- Email: [email protected]
Lung damage from exposure to the fields of an electrohydraulic lithotripter
Ultrasound in Med. & Biol. Vol. 16, No. 7, pp. 675-679, 1990 Prinled in the U.S.A.
0301-5629/90 $3.00 + .00 © 1990 Pergamon Press plc
OOriginal Contribution LUNG DAMAGE FROM EXPOSURE TO THE FIELDS OF AN ELECTROHYDRAULIC LITHOTRIPTER C. HARTMAN, t~ S. Z. CHILD, t~ R. MAYER, S§ E. SCHENK ~:11a n d E. L. CARSTENSEN t$¶ Departments of*Electrical Engineering, ~Urology, ttPathologyand ~Biophysicsand the *Rochester Center for Biomedical Ultrasound, The University of Rochester, Rochester, NY 14627 (Receiw,d 15 February 1990; in fmalJorm 2l May 1990) Abstract--Threshold pressures for hemorrhage in mouse lung exposed to the fields of an electrohydraulic lithotripter appear to be less than 2 M P a with as few as 10 pulses and with severe damage occurring at levels between 5 and 6 M P a . This is very much smaller than the fields required to fragment kidney and gallstones and smaller than the thresholds for damage to kidney tissues. Fetal lung, in contrast, did not show signs of damage at 20 MPa. The lower sensitivity of fetal lung is consistent with a cavitation-related mechanism for lung damage by shock waves. Since the pressures in these exposures are almost entirely positive, it suggests that the value of negative pressures as predictors of the behavior of gas bodies in tissues should be reconsidered.
Key Words: Lithotripsy, Cavitation, Shock waves, Lung hemorrhage, Mouse lung.
INTRODUCTION
kidney studies, was a spherically diverging, almost entirely positive wave. The source was a 9-French probe of a Wolf Model 2137.50 Electrohydraulic Lithotripter (Richard Wolf GMBH, Postfach 4D, D-7134 Knittlingen, West Germany). The fields and behavior of the primary bubble, which is generated upon ignition of the spark, have been characterized by Campbell et al. (1990). The shock wavefront is a nearly uniform hemisphere. Each spark discharge produces two, spherically diverging shock waves, the first at ignition and the second approximately I ms later from the rebound of the gas bubble. Pressure waveforms as measured with a Marconi (GEC Marconi Research, Chelmsford, UK) bilaminar, PVDF membrane hydrophone are shown in Fig. 1. Pressures are almost entirely positive, i.e., the negative pressures involved in the exposure were immeasurably small (Campbell et al. 1990). The amplitudes of the two pressure pulses are roughly equal. For the animal exposures, shocks were administered at the rate of approximately 1 per second. Although there is some variability from spark to spark, the pressure amplitude is typically 3 MPa at 3 cm from the source. With a spherically diverging wave, the pressure is inversely proportional to distance from the source. Using this relationship, free field measurements of the pressure field at 3 cm were used to select the probe-animal distance necessary to achieve the desired exposure levels. To minimize the possibility of complications
Lung damage has been observed to result from exposure of experimental animals to fields of an extracorporeal shock wave lithotripter (ESWL). Delius et al. (1987) used a Dornier HM-II in studies with beagle and boxer dogs. This device is characterized by a sharp positive pressure spike followed by a somewhat more slowly varying negative pressure which is typically less than half the positive pressure amplitude. The results of their studies were mixed but hemorrhages occurred frequently at positive pressures of 10 MPa (negative pressures ~ 3-6 MPa) and rarely at positive pressures of 2 MPa (negative pressures of 1 MPa). In the course of a study of extravasation in mouse kidneys using a spherically diverging, sparkgenerated shock wave (Mayer et al. 1990), we noted lung hemorrhage and, in fact, were limited in some aspects of our protocol because of these "side effects." When we attempted to study the influence of pulse number on the kidney, we rarely could administer more than 100 shocks before the animal succumbed from massive lung damage. We report here a direct study of the threshold for hemorrhage in mouse lung. EXPERIMENTAL PROCEDURE In contrast with reports from other laboratories, the shock wave source used here, as in our earlier 675
676
Ultrasound in Medicine and Biology
Volume16, Number 7, 1990
(a)
(b)
Fig. 1. Shock waveforms of the Wolf Model 2137.50 Electrohydraulic Lithotripter. (a) Primary shock which is generated at ignition. (b) The second shock which is initiated by rebound of the bubble that is created by the spark. Sweep rate, 1 us/div. These observations were made with a Marconi, bilaminar PVDF membrane hydrophone at a distance of 3 cm from the spark source. Pressures at this distance are typically 3 MPa and vary in inverse proportion to the distance from the source. from reflections at the surface o f the water, a waterfilled cylindrical container with an acoustically transparent b o t t o m was placed at the surface of the water in the space between probe and mouse (Fig. 2). Hydrophone m e a s u r e m e n t s showed that this effectively e l i m i n a t e d surface reflections. In this study, the p r o b e - a n i m a l separations were so large (>5 cm at threshold) that there is no possibility of direct contact o f the primary bubble with the surface o f the mouse during the bubble's active lifetime.
Adult lung Male C 3 H mice, a p p r o x i m a t e l y 7 weeks old, were o b t a i n e d f r o m J a c k s o n L a b s (Bar H a r b o r , Maine), fed ad libitum and cared for in accordance with institutional guidelines on animal care. Before exposure, the mice were anesthetized with K e t a m i n e (200 mg/kg) and R o m p u n (10 mg/kg). Their backs were shaved with clippers and the remaining fur was removed with Neet ® to minimize the likelihood of e n t r a p m e n t of air at the skin-water interface. The
Water
Water
Window
Rubber
Fig. 2. Diagram of exposure arrangement. A water-filled cylinder with a film window on the bottom was placed between probe and animal to minimize reflections from the surface.
Lung damage from an electrohydraulic lithotripter • C. HARTMAN et al.
animals were positioned on a rubber flame with the dorsal surface toward the source. The frame, in turn, was supported on a three-way positioner that allowed stable, precise positioning of the mouse with respect to the spark source. The probe was centered over the thorax of the animal. The pressures reported here are the maximum values at the surface of the animal. However, the source to skin distances were large enough that at threshold the entire lung of the animal received uniform exposure within 3-4%. Attenuation by the thoracic wall was measured by comparing primary shock waves as measured with a PVDF needle hydrophone in the free field with the waves received when the skin-covered thoracic wall of a mouse (viscera removed) was placed around the end of the hydrophone. The attenuation was less than the pulse-to-pulse variability in wave amplitude, i.e., <1 db. Immediately following exposure, the lungs were exposed surgically and examined for gross signs of hemorrhage. The animals were killed by cervical dislocation. In this study, the most sensitive, consistent indicator of damage to the lung was a collection of blood at the base of the organ. Selected lungs were fixed in 10% neutral buffered formaldehyde, paraffin embedded, sectioned longitudinally at 5 microns, and stained with hematoxylin and eosin and the Gomori trichrome stain for connective tissues.
Fetal lung A special experiment was conducted to confirm the postulate that lung sensitivity is correlated with the presence of air in that tissue. For this test, mouse fetuses on the 18th day of gestation were exposed to the same shockwave source. In these experiments, the pregnant female was anesthetized and mounted for exposure. The lithotripter probe was placed directly over the uterus at a distance such that the shock wave pressures experienced by the fetuses were approximately 20 MPa, an order of magnitude greater than the threshold for damage to adult lung. Immediately following exposure, the dam was sacrificed by cervical dislocation. Fetuses were removed and examined with particular attention to the lung. EXPERIMENTAL RESULTS The threshold data for hemorrhage in adult mouse lung from exposure to spherically diverging shock waves are presented in Fig. 3. The ordinate is the peak positive pressure at the surface of the animal. Most of the observations were made with a series of 10 sparks (10 double shock waves). Observations
677
• Positive I [] Negative
5 4 3
mmD
•••••
DDD
In DD~BD -D m-u
2
~
u~
aE~rn
I1•
smm
DDD
0.5 5
10
20
Pulse Number
Fig. 3. Threshold determinations for damage to mouse lung exposed to the shock fields described in Fig. 1. Each point represents observations of lung hemorrhage for a single experimental animal. The lateral spread of data points is for clarity in the presentation. For example, 12 animals were exposed to l0 double pulses at a pressure of 2 MPa.
were also conducted with 5 and 20 exposures. Each point in Fig. 3 represents the results from a single experimental animal. No attempt is made in that figure to describe the degree of damage. However, in a qualitative sense, there was a sharp increase in the extent of damage as the pressure increased. As the data show, for the particular indicator of damage used in these studies, a rather clear threshold is shown that decreases monotonically from approximately 2.5 MPa with 5 double pulses to less than 1.5 MPa with 20 pulses. Microscopic examination showed intra-alveolar hemorrhage in the basal areas, usually most severe subpleurally along the dorsal aspect (Fig. 4). Sections stained for connective tissues showed extensive disruption of alveolar walls. Bronchi and peribronchial blood vessels showed no apparent structural changes, although extravasated red blood cells were frequently present within bronchial lumina. Exposures of pregnant mice to 10 double shocks at a peak positive pressure of 20 MPa at the surface of the skin over the fetuses resulted in marked damage to the dams, e.g., skin damage directly below the spark source and over the fetuses, intestinal and lung hemorrhages, the latter in spite of the fact that the pressures at the adult lung were somewhat less than at the site of the fetus. In contrast, the fetuses were remarkably free of signs of gross injury. Out of 50 fetuses in five litters examined, only 2 showed any sign of lung damage. Another 4 fetuses showed evidence of hemorrhage at other locations.
678
Ultrasound in Medicine and Biology
Volume 16, Number 7, 1990
(a)
(b) Fig. 4. (a) Normal mouse lung. Hematoxylin and eosin stain, x 100. (b) Basal portion of mouse lung exposed to the shock field shows extensive intra-alveolar hemorrhage most prominently along the dorsal subpleural aspect. Hematoxylin and eosin, X40.
DISCUSSION In spite o f some limitations, the present experimental system has several advantages over use of focused sound fields and large animals. With focused
fields, in-situ acoustic pressure estimates are subject to considerable error. With large animals, it is necessary to m a k e corrections for attenuation through the tissues and because o f spatial variations in the focused field it is difficult after exposure to identify a
Lung damage from an electrohydraulic lithotripter • C. HARTMAN et al.
given pressure with a given anatomical site. With the small animals used here, attenuation is negligible and, with the possible exception of the highest frequencies in the shock fields, the pressures in the animal are essentially the same as they are in water at the same distance from the source. As discussed below, the relatively simple, largely positive pressures of the source used here have p r o d u c e d i n f o r m a t i o n that would have been lost with focused sources. Most important, with the point source used in these studies, the field is u n i f o r m over the forward hemisphere. Thus, in-situ pressures in the mouse depend only on the accuracy of pressure measurements in water and the source to target distances. By the nature o f the endpoint chosen for observation, the value 1.5 MPa with 20 double pulses must be taken as an upper limit to the real threshold value. A m o r e sensitive i n d i c a t o r might yield a smaller threshold value. Furthermore, it is possible that the apparent dependence o f threshold on pulse n u m b e r results, in part, simply from the sensitivity of our assay. For clinical purposes, it would be useful to know a pressure level below which the lung could be exposed indefinitely, or at least to several thousand pulses, without damage. F r o m the physical nature o f cavitation, it is reasonable to believe that such a threshold exists. Other experimental designs will be required to determine the actual value. It would have been difficult in this study to have used pressure levels less than 1 MPa. By way of caution, it must be noted that pressures above 2 MPa extend far outside, in front of and beyond the focal region ofextracorporeal shock wave lithotripters. With regard to the threshold values themselves, it appears that damage to these particularly sensitive tissues occurs at pressure levels much smaller than those required to fragment stones. That the threshold for damage to fetal lung is at least 10 times greater than for adult lung is consistent with the conventional assumption that the lung gets its sensitivity by virtue of the presence of air bodies which are potential sites for cavitation-related activity. With that in mind, it is particularly noteworthy that the threshold
679
for hemorrhage in mouse kidney is within a factor of two or three of the thresholds found in adult mouse lung (Mayer et al. 1990). For 10 pressure pulses, the mouse kidney threshold is of the order of 3-5 MPa, while lung is approximately 1.7 MPa. In spite of the qualitatively different pressure fields and target tissues used here and those used in the pioneering studies by Delius et al. (1987), our threshold values for lung damage are near the lower limits of the 2-10 MPa (positive pressure) limits set by those workers in their study with extracorporeal shockwave lithotripsy in dogs. Their exposures involved both positive and negative pressures whereas, in our case, negative pressures are immeasurably small. It would be tempting to conclude from these observations that positive rather than negative pressures were key parameters in predicting tissue damage. This would be contrary to conventional wisdom regarding the cavitation mechanism. There are other observations which have raised this question (Carstensen et al. 1990; Hartman et al. 1990; Mayer et al. 1990). It remains to be determined in future studies whether a classical cavitation mechanism can explain the response of tissues like lung to lithotripter and pulsed ultrasound fields. Acknowledgments--This work was supported in part by U.S.P.H.S. Grant No. DK39796.
REFERENCES Campbell, D. S.; Flynn, H. G.; Blackstock, D. T.; Linke, C.; Carstensen, E. L. The acoustic fields of the Wolf electrohydraulic lithotripter. J. Litho. Stone Dis. [ 1990]. Carstensen, E. L.; Campbell, D. S.; Hoffman, D.; Child, S. Z.; Aymr-Bellegarda, E. J. Killing of Drosophila larvae by the fields of an electrohydraulic lithotripter. Ultrasound Med. Biol. 16(7):687-698; 1990. Delius, M.; Enders, G.; Heine, G.; Stark, J.; Remberger, K.; Brendel, W. Biological effects of shock waves: Lung hemorrhage by shock waves in dogs--Pressure dependence. Ultrasound Med. Biol. 13:61-67; 1987. Hartman, C.; Cox, C. A.; Brewer, L.; Child, S. Z.; Cox, C. F.; Carstensen, E. L. Effects oflithotripter fields on development of chick embryos. Ultrasound Med. Biol. 16(6):581-585; 1990. Mayer, R.; Schenk, E.; Child, S.; Norton, S.; Cox, C. A.; Hartman, C.; Cox, C. F.; Carstensen, E. L. Pressure threshold for shockwave induced renal hemorrhage. J. Urol. [1990].
0301-5629/90 $3.00 + .00 © 1990 Pergamon Press plc
OOriginal Contribution LUNG DAMAGE FROM EXPOSURE TO THE FIELDS OF AN ELECTROHYDRAULIC LITHOTRIPTER C. HARTMAN, t~ S. Z. CHILD, t~ R. MAYER, S§ E. SCHENK ~:11a n d E. L. CARSTENSEN t$¶ Departments of*Electrical Engineering, ~Urology, ttPathologyand ~Biophysicsand the *Rochester Center for Biomedical Ultrasound, The University of Rochester, Rochester, NY 14627 (Receiw,d 15 February 1990; in fmalJorm 2l May 1990) Abstract--Threshold pressures for hemorrhage in mouse lung exposed to the fields of an electrohydraulic lithotripter appear to be less than 2 M P a with as few as 10 pulses and with severe damage occurring at levels between 5 and 6 M P a . This is very much smaller than the fields required to fragment kidney and gallstones and smaller than the thresholds for damage to kidney tissues. Fetal lung, in contrast, did not show signs of damage at 20 MPa. The lower sensitivity of fetal lung is consistent with a cavitation-related mechanism for lung damage by shock waves. Since the pressures in these exposures are almost entirely positive, it suggests that the value of negative pressures as predictors of the behavior of gas bodies in tissues should be reconsidered.
Key Words: Lithotripsy, Cavitation, Shock waves, Lung hemorrhage, Mouse lung.
INTRODUCTION
kidney studies, was a spherically diverging, almost entirely positive wave. The source was a 9-French probe of a Wolf Model 2137.50 Electrohydraulic Lithotripter (Richard Wolf GMBH, Postfach 4D, D-7134 Knittlingen, West Germany). The fields and behavior of the primary bubble, which is generated upon ignition of the spark, have been characterized by Campbell et al. (1990). The shock wavefront is a nearly uniform hemisphere. Each spark discharge produces two, spherically diverging shock waves, the first at ignition and the second approximately I ms later from the rebound of the gas bubble. Pressure waveforms as measured with a Marconi (GEC Marconi Research, Chelmsford, UK) bilaminar, PVDF membrane hydrophone are shown in Fig. 1. Pressures are almost entirely positive, i.e., the negative pressures involved in the exposure were immeasurably small (Campbell et al. 1990). The amplitudes of the two pressure pulses are roughly equal. For the animal exposures, shocks were administered at the rate of approximately 1 per second. Although there is some variability from spark to spark, the pressure amplitude is typically 3 MPa at 3 cm from the source. With a spherically diverging wave, the pressure is inversely proportional to distance from the source. Using this relationship, free field measurements of the pressure field at 3 cm were used to select the probe-animal distance necessary to achieve the desired exposure levels. To minimize the possibility of complications
Lung damage has been observed to result from exposure of experimental animals to fields of an extracorporeal shock wave lithotripter (ESWL). Delius et al. (1987) used a Dornier HM-II in studies with beagle and boxer dogs. This device is characterized by a sharp positive pressure spike followed by a somewhat more slowly varying negative pressure which is typically less than half the positive pressure amplitude. The results of their studies were mixed but hemorrhages occurred frequently at positive pressures of 10 MPa (negative pressures ~ 3-6 MPa) and rarely at positive pressures of 2 MPa (negative pressures of 1 MPa). In the course of a study of extravasation in mouse kidneys using a spherically diverging, sparkgenerated shock wave (Mayer et al. 1990), we noted lung hemorrhage and, in fact, were limited in some aspects of our protocol because of these "side effects." When we attempted to study the influence of pulse number on the kidney, we rarely could administer more than 100 shocks before the animal succumbed from massive lung damage. We report here a direct study of the threshold for hemorrhage in mouse lung. EXPERIMENTAL PROCEDURE In contrast with reports from other laboratories, the shock wave source used here, as in our earlier 675
676
Ultrasound in Medicine and Biology
Volume16, Number 7, 1990
(a)
(b)
Fig. 1. Shock waveforms of the Wolf Model 2137.50 Electrohydraulic Lithotripter. (a) Primary shock which is generated at ignition. (b) The second shock which is initiated by rebound of the bubble that is created by the spark. Sweep rate, 1 us/div. These observations were made with a Marconi, bilaminar PVDF membrane hydrophone at a distance of 3 cm from the spark source. Pressures at this distance are typically 3 MPa and vary in inverse proportion to the distance from the source. from reflections at the surface o f the water, a waterfilled cylindrical container with an acoustically transparent b o t t o m was placed at the surface of the water in the space between probe and mouse (Fig. 2). Hydrophone m e a s u r e m e n t s showed that this effectively e l i m i n a t e d surface reflections. In this study, the p r o b e - a n i m a l separations were so large (>5 cm at threshold) that there is no possibility of direct contact o f the primary bubble with the surface o f the mouse during the bubble's active lifetime.
Adult lung Male C 3 H mice, a p p r o x i m a t e l y 7 weeks old, were o b t a i n e d f r o m J a c k s o n L a b s (Bar H a r b o r , Maine), fed ad libitum and cared for in accordance with institutional guidelines on animal care. Before exposure, the mice were anesthetized with K e t a m i n e (200 mg/kg) and R o m p u n (10 mg/kg). Their backs were shaved with clippers and the remaining fur was removed with Neet ® to minimize the likelihood of e n t r a p m e n t of air at the skin-water interface. The
Water
Water
Window
Rubber
Fig. 2. Diagram of exposure arrangement. A water-filled cylinder with a film window on the bottom was placed between probe and animal to minimize reflections from the surface.
Lung damage from an electrohydraulic lithotripter • C. HARTMAN et al.
animals were positioned on a rubber flame with the dorsal surface toward the source. The frame, in turn, was supported on a three-way positioner that allowed stable, precise positioning of the mouse with respect to the spark source. The probe was centered over the thorax of the animal. The pressures reported here are the maximum values at the surface of the animal. However, the source to skin distances were large enough that at threshold the entire lung of the animal received uniform exposure within 3-4%. Attenuation by the thoracic wall was measured by comparing primary shock waves as measured with a PVDF needle hydrophone in the free field with the waves received when the skin-covered thoracic wall of a mouse (viscera removed) was placed around the end of the hydrophone. The attenuation was less than the pulse-to-pulse variability in wave amplitude, i.e., <1 db. Immediately following exposure, the lungs were exposed surgically and examined for gross signs of hemorrhage. The animals were killed by cervical dislocation. In this study, the most sensitive, consistent indicator of damage to the lung was a collection of blood at the base of the organ. Selected lungs were fixed in 10% neutral buffered formaldehyde, paraffin embedded, sectioned longitudinally at 5 microns, and stained with hematoxylin and eosin and the Gomori trichrome stain for connective tissues.
Fetal lung A special experiment was conducted to confirm the postulate that lung sensitivity is correlated with the presence of air in that tissue. For this test, mouse fetuses on the 18th day of gestation were exposed to the same shockwave source. In these experiments, the pregnant female was anesthetized and mounted for exposure. The lithotripter probe was placed directly over the uterus at a distance such that the shock wave pressures experienced by the fetuses were approximately 20 MPa, an order of magnitude greater than the threshold for damage to adult lung. Immediately following exposure, the dam was sacrificed by cervical dislocation. Fetuses were removed and examined with particular attention to the lung. EXPERIMENTAL RESULTS The threshold data for hemorrhage in adult mouse lung from exposure to spherically diverging shock waves are presented in Fig. 3. The ordinate is the peak positive pressure at the surface of the animal. Most of the observations were made with a series of 10 sparks (10 double shock waves). Observations
677
• Positive I [] Negative
5 4 3
mmD
•••••
DDD
In DD~BD -D m-u
2
~
u~
aE~rn
I1•
smm
DDD
0.5 5
10
20
Pulse Number
Fig. 3. Threshold determinations for damage to mouse lung exposed to the shock fields described in Fig. 1. Each point represents observations of lung hemorrhage for a single experimental animal. The lateral spread of data points is for clarity in the presentation. For example, 12 animals were exposed to l0 double pulses at a pressure of 2 MPa.
were also conducted with 5 and 20 exposures. Each point in Fig. 3 represents the results from a single experimental animal. No attempt is made in that figure to describe the degree of damage. However, in a qualitative sense, there was a sharp increase in the extent of damage as the pressure increased. As the data show, for the particular indicator of damage used in these studies, a rather clear threshold is shown that decreases monotonically from approximately 2.5 MPa with 5 double pulses to less than 1.5 MPa with 20 pulses. Microscopic examination showed intra-alveolar hemorrhage in the basal areas, usually most severe subpleurally along the dorsal aspect (Fig. 4). Sections stained for connective tissues showed extensive disruption of alveolar walls. Bronchi and peribronchial blood vessels showed no apparent structural changes, although extravasated red blood cells were frequently present within bronchial lumina. Exposures of pregnant mice to 10 double shocks at a peak positive pressure of 20 MPa at the surface of the skin over the fetuses resulted in marked damage to the dams, e.g., skin damage directly below the spark source and over the fetuses, intestinal and lung hemorrhages, the latter in spite of the fact that the pressures at the adult lung were somewhat less than at the site of the fetus. In contrast, the fetuses were remarkably free of signs of gross injury. Out of 50 fetuses in five litters examined, only 2 showed any sign of lung damage. Another 4 fetuses showed evidence of hemorrhage at other locations.
678
Ultrasound in Medicine and Biology
Volume 16, Number 7, 1990
(a)
(b) Fig. 4. (a) Normal mouse lung. Hematoxylin and eosin stain, x 100. (b) Basal portion of mouse lung exposed to the shock field shows extensive intra-alveolar hemorrhage most prominently along the dorsal subpleural aspect. Hematoxylin and eosin, X40.
DISCUSSION In spite o f some limitations, the present experimental system has several advantages over use of focused sound fields and large animals. With focused
fields, in-situ acoustic pressure estimates are subject to considerable error. With large animals, it is necessary to m a k e corrections for attenuation through the tissues and because o f spatial variations in the focused field it is difficult after exposure to identify a
Lung damage from an electrohydraulic lithotripter • C. HARTMAN et al.
given pressure with a given anatomical site. With the small animals used here, attenuation is negligible and, with the possible exception of the highest frequencies in the shock fields, the pressures in the animal are essentially the same as they are in water at the same distance from the source. As discussed below, the relatively simple, largely positive pressures of the source used here have p r o d u c e d i n f o r m a t i o n that would have been lost with focused sources. Most important, with the point source used in these studies, the field is u n i f o r m over the forward hemisphere. Thus, in-situ pressures in the mouse depend only on the accuracy of pressure measurements in water and the source to target distances. By the nature o f the endpoint chosen for observation, the value 1.5 MPa with 20 double pulses must be taken as an upper limit to the real threshold value. A m o r e sensitive i n d i c a t o r might yield a smaller threshold value. Furthermore, it is possible that the apparent dependence o f threshold on pulse n u m b e r results, in part, simply from the sensitivity of our assay. For clinical purposes, it would be useful to know a pressure level below which the lung could be exposed indefinitely, or at least to several thousand pulses, without damage. F r o m the physical nature o f cavitation, it is reasonable to believe that such a threshold exists. Other experimental designs will be required to determine the actual value. It would have been difficult in this study to have used pressure levels less than 1 MPa. By way of caution, it must be noted that pressures above 2 MPa extend far outside, in front of and beyond the focal region ofextracorporeal shock wave lithotripters. With regard to the threshold values themselves, it appears that damage to these particularly sensitive tissues occurs at pressure levels much smaller than those required to fragment stones. That the threshold for damage to fetal lung is at least 10 times greater than for adult lung is consistent with the conventional assumption that the lung gets its sensitivity by virtue of the presence of air bodies which are potential sites for cavitation-related activity. With that in mind, it is particularly noteworthy that the threshold
679
for hemorrhage in mouse kidney is within a factor of two or three of the thresholds found in adult mouse lung (Mayer et al. 1990). For 10 pressure pulses, the mouse kidney threshold is of the order of 3-5 MPa, while lung is approximately 1.7 MPa. In spite of the qualitatively different pressure fields and target tissues used here and those used in the pioneering studies by Delius et al. (1987), our threshold values for lung damage are near the lower limits of the 2-10 MPa (positive pressure) limits set by those workers in their study with extracorporeal shockwave lithotripsy in dogs. Their exposures involved both positive and negative pressures whereas, in our case, negative pressures are immeasurably small. It would be tempting to conclude from these observations that positive rather than negative pressures were key parameters in predicting tissue damage. This would be contrary to conventional wisdom regarding the cavitation mechanism. There are other observations which have raised this question (Carstensen et al. 1990; Hartman et al. 1990; Mayer et al. 1990). It remains to be determined in future studies whether a classical cavitation mechanism can explain the response of tissues like lung to lithotripter and pulsed ultrasound fields. Acknowledgments--This work was supported in part by U.S.P.H.S. Grant No. DK39796.
REFERENCES Campbell, D. S.; Flynn, H. G.; Blackstock, D. T.; Linke, C.; Carstensen, E. L. The acoustic fields of the Wolf electrohydraulic lithotripter. J. Litho. Stone Dis. [ 1990]. Carstensen, E. L.; Campbell, D. S.; Hoffman, D.; Child, S. Z.; Aymr-Bellegarda, E. J. Killing of Drosophila larvae by the fields of an electrohydraulic lithotripter. Ultrasound Med. Biol. 16(7):687-698; 1990. Delius, M.; Enders, G.; Heine, G.; Stark, J.; Remberger, K.; Brendel, W. Biological effects of shock waves: Lung hemorrhage by shock waves in dogs--Pressure dependence. Ultrasound Med. Biol. 13:61-67; 1987. Hartman, C.; Cox, C. A.; Brewer, L.; Child, S. Z.; Cox, C. F.; Carstensen, E. L. Effects oflithotripter fields on development of chick embryos. Ultrasound Med. Biol. 16(6):581-585; 1990. Mayer, R.; Schenk, E.; Child, S.; Norton, S.; Cox, C. A.; Hartman, C.; Cox, C. F.; Carstensen, E. L. Pressure threshold for shockwave induced renal hemorrhage. J. Urol. [1990].