T H E R A P E U T I C ULTRASOUND IN T H E PRODUCTION O F OCULAR LESIONS D. J A C K S O N C O L E M A N , M.D., F R E D E R I C L. L I Z Z I , E N G . S C . D . A N D F R E D E R I C K A. J A K O B I E C , M.D.
New York, New York
The therapeutic application of con trolled ocular damage has gained wide spread use in the forms of xenon arc photocoagulation, laser photocoagulation, diathermy, and cryopexy. Each of these therapeutic modalities has specific features that make it more applicable to some situations than to others. All of these methods require precise control to produce the desired therapeutic response. A separate modality, therapeutic ultra sound, produces a similar type of tissue damage. Until recently, however, data on tissue effects and methods of control were so inadequate that such a technique could not be considered for ophthalmic treat ment. Through our laboratory studies, we have addressed these inadequacies and developed a therapeutic ultrasound sys tem for clinical use. The advantage of therapeutic ultra sound is its unique ability to produce lesions independent of the optical tissue absorption properties that limit light photocoagulation systems. This permits control of lesion placement and severity chiefly by system factors rather than by the pigmentation of the target tissue. The " b u r n " spot can be positioned in virtually any area of tissue (within the focal limits of the system), thus allowing treatment of ocular and orbital abnormalities that pre viously could be treated only after surgiFrom the Department of Ophthalmology (Drs. Coleman and Jakobiec), College of Physicians and Surgeons, Columbia University, and Riverside Re search Institute (Dr. Lizzi) New York, New York. This research was supported in part by Public Health Service Grants EY-01480 and EY-01218 from the National Eye Institute. Reprint requests to D. Jackson Coleman, M.D., E. S. Harkness Eye Institute, 635 West 165th St., New York, NY 10032.
cal exposure. The ultrasonic system af fords two advantages: it may be applied externally, as is photocoagulation; and, as is the case with radiotherapy techniques, it does not require direct visualization for the treatment of lesions. For example, patients with retinitis proliferans ob scured by opaque media or tumors ob scured by hemorrhage are candidates for ultrasonic treatment. Experience in the production of ocular lesions was described by Donn 1 ; Purnell, Sokollu, and Holasek 2 ; Torchia, Purnell, and Sokollu 3 ; Coleman and associates 4 ; and Lizzi, Burt, and Coleman. 5 These investigators attemped to produce ocular lesions chiefly in the vitreous, lens, retina, and choroid and their primary aim was to identify the minimal exposure levels re quired to produce clinically useful le sions while not damaging the adjacent ocular tissues. Difficulties in determining the proper productive energy levels and in precisely positioning the therapeutic beam had previously prevented clinical utilization of ultrasound for ocular therapy. We have studied the biologic effects of ultrasound in ocular tissues for the past five years. Meticulous control of energies required for production of corneal, lentic ular, and chorioretinal lesions has yielded threshold data on experimental animals. 6 A graph summarizing these results, as well as those salient findings of other laboratories, is shown (Fig. 1). The ener gy levels required to produce a clinically useful therapeutic lesion generally exceed energy levels of 50 W/cm 2 ; conversely, if energy levels are kept below this level in surrounding tissues, then adjacent tissue damage can be avoided. The mechanisms that cause tissue dam-
AMERICAN JOURNAL OF OPHTHALMOLOGY 86:185-192, 1978
185
186
AMERICAN JOURNAL O F OPHTHAMOLOGY ULTRASONIC INSONIFICATION OF OCULAR TISSUES o - ZISKIN
ET AL
b - BAUM c - PURNELL AND SOKOlLU d - L I 2 Z I AND COLEMAN
CATARACTSI3 5 M H z , c )
NO EFFECT 11MHz, b)
NO EFFECT 19.5 MHz, a 1
-X-
Fig. 1 (Coleman, Lizzi, and Jakobiec). Available data are superimposed to indicate ultrasonic thresh old levels for damage found in ocular tissues. The dotted line indicates the energy levels below which no tissue damage has been noted and above which the damage thresholds appear to fall.
A U G U S T , 1978
energy. The choroidal blanching proba bly reflects an alteration of blood flow that may cause the thermal reaction by reduction of heat removal. The curves for ultrasonically induced lesions compared with data derived by Ham 7 for the production of chorioretinal lesions by use of laser and xenon arc irradiation are shown (Fig. 2). The opti cally derived data used a focal spot of 0.8 mm; this is comparable to the 0.4 mmbeam we used. The similarly shaped curves suggest shared damage mecha nisms. Optical energy is absorbed chiefly in the pigment epithelium of the retina, whereas ultrasonic energy is absorbed more diffusely across the retina and choroid. Since the optic absorption coeffi cients of the pigment epithelium are extremly high as compared to that of the average ultrasonic coefficient, it is not surprising that larger amounts of ultra sonic energy are needed for similar lesion
2000
age are not completely understood, but present data suggest several phenomena may be involved in lesion formation. In chorioretinal lesions, a transitory choroi dal blanching that is easily seen in "subthreshold" lesions is consistent with a radiation pressure mechanism first sug gested by Purnell, Sokollu, and Holasek. 2 The blanching is probably produced by streaming forces that arise from absorptive gradients and radiation forces within the vitreous pathway preceding the lesion. The second form of lesion, a "threshold" lesion, is characterized by choroidal blanching, as well as an overlying edema of the retina, which ensues within 24 hours after exposure to insonification; these lesions exhibit small permanent pigment epithelial and chorioretinal ef fects. The permanent damage lesion is considered to be caused by thermal phe nomena arising from absorbed ultrasonic
1000
&
WHITE LIGHT ( \ < 9 5 0 m M )
•
ULTRASOUND ( 9 . 6 MHz)
TOO
i I
NOTE: LIGHT INTENSITY IS MULTIPLIED BY 25
400
200
100
:
^ 9 . 8 MHz ULTRASOUND (0.4mm »pol/»il«)
/
TO
- - -
^ W H I T E LIGHT (0.8 mm i p o t / i i i e )
40
20
10
, EXPOSURE MINIMAL
DURATION
(tec)
FUNDUSCOPICALLY VISIBLE CHORIORETINAL LESIONS
Fig. 2 (Coleman, Lizzi, and Jakobiec). The dam age threshold curve found with ultrasound is com pared with a damage threshold curve for chorioreti nal lesions produced by white light. The optic absorption coefficient of the pigment epithelium (white light damage) is extremely high compared to that of the average ultrasonic coefficient. Allowing for these differences in absorption thresholds, for light as compared to ultrasound, the curves are remarkably similar. The similarity of the shapes of these curves suggest a similar damage mechanism for both light and ultrasonic energy.
VOL. 86, NO. 2
THERAPEUTIC ULTRASOUND
formation. Consequently, the intensity levels for ultrasonic energy must be mul tiplied by a factor of 25 to obtain a com parable curve, but the time durations have not been scaled. With the understanding of the energy re quired to produce toxic ultrasonic effects and an understanding of the transducer design factors required to produce both the desired therapeutic effect while spar ing adjacent tissue from damage, a thera peutic system can be considered for oph thalmic application. We have developed a therapeutic sys tem intended for use in conjunction with a previously described diagnostic ultra sound visualization technique. 8 The diag nostic system uses low-energy pidsed ul trasound that produces no known tissue damage. The intensity levels of the diag nostic system are smaller than those of the therapeutic system by a factor of 10 5 , and therefore do not influence the intend ed therapeutic effects. The position of the focal zone of the therapeutic transducer appears on the diagnostic B-scan or Ascan display, or both, as an intensified area that allows manual positioning of the therapeutic focus within the targeted zone. Once the desired position is ob tained by observing the diagnostic dis play, the burn is then produced in milli second intervals in much the same manner that a laser treatment is applied. Extensive evaluation of ultrasonic lesions produced in experimental animals indi cates a unique potential for clinical appli cation in various disease states.
187
Fig. 3 (Coleman, Lizzi, and Jakobiee). The thera peutic transducer is shown with a central aperture cut out for placement of a paraxial diagnostic trans ducer. The therapeutic transducer is a focused ce ramic with a focal length of 9.0 cm and operates at a frequency of 9.8 MHz.
element (Fig. 3). A schlieren photograph of the focal zone of the therapeutic trans ducer is shown (Fig. 4). The energy levels used produce an ocu lar burn only in the focal zone. Levels ranging from 500 to 3,000 W/cm 2 in this focal zone have been used to produce lesions of the type reported here. Sur rounding regions of lower ultrasonic en ergy intensity have been found to be safe
METHOD
We designed a spherical focused trans ducer operated at a frequency of 9.8 MHz to produce a focal spot size of 0.4 mm in diameter and 4 mm in length at a distance of 9.0 cm from the transducer surface. We removed the center of the transducer to allow a diagnostic transducer to be placed paraxial with the surrounding therapeutic
Fig. 4 (Coleman, Lizzi, and Jakobiee). Schlierien photography of the therapeutic transducer beam used to produce the lesions shown in Figures 6 through 10. The focal zone is sharply angled to provide a burn of approximately 0.4 mm in diameter and 4 mm in length. The energy in surrounding areas of tissue is weak enough to produce no known damage.
188
AMERICAN JOURNAL OF OPHTHAMOLOGY
in extensive experimental studies. 9 We have demonstrated the manner in which the diagnostic and therapeutic systems are combined (Fig. 5). Both systems are manually operated, so that positioning of the beam and application of the treatment are under the physician's control. B-scan display is made by sweeping the diagnos tic transducer across the eye and the image is stored on the oscilloscope. The intensified area representing the focal zone of the therapeutic transducer is then positioned until it corresponds to the area of disease to be treated. At this point, the stored image is erased and direct kinetic diagnostic scanning is used to allow the final localization of the area of disease. In this way, the location of the therapeutic beam can be continuously observed. In clear media, vitreal, retinal, and choroidal lesions can be seen as they form. The appearance of a white area of retina visu ally resembles the lesion produced by laser or xenon photocoagulation. This system has been tested in a series of 80 albino and Dutch pigmented rab bits. The therapeutic transducer has been used to produce lesions of the cornea, lens, and the choroid, retina, sclera com plex.
AUGUST, 1978 RESULTS
Chorioretinal lesions have been an area of chief interest because the use of ultra sound in the production of such lesions is potentially of great clinical value, and the lesions can be compared with lesions produced by other modalities. Exper iments have been performed to evaluate the dimensions and nature of lesions that are created at specific ultrasonic expo sures (Figs. 6 and 7). Histologic examination of acute suprathreshold lesions demonstrated damage to the outer retina, disruption and necro sis of the pigment epithelium, subretinal exudate, mild choroidal hemorrhage, col lections of polymorphonuclear leukocytes in and around the choroidal vessels, and loss of subadjacent scleral nuclei. The trichrome stain revealed that the sclera was also altered in its staining properties, thus suggesting a physical modification of the collagen (Fig. 8). In contrast to laser-treated eyes, where most of the dam age is restricted to the level of the pig ment epithelial cells or other pigmented tissue, all layers of the ultrasonically treated eyes display some damage. The damage, however, is variable for the dif ferent tunics. The retina in mild supra-
Fig. 5 (Coleman, Lizzi, and Jakobiec). The combined therapeuticdiagnostic system as designed for clinical evaluation. The diagnostic system provides the azimuthal and axial localization, regardless of me dia clarity for positioning the treat ment zone of the therapeutic trans ducer.
Fig. 6 (Coleman, Lizzi, and Jakobiec). Fundus photograph of an albino rabbit showing a suprathreshold lesion 24 hours after insonification.
Fig. 7 (Coleman, Lizzi, and Jakobiec). Fundus photograph of an albino rabbit showing a high intensity lesion seven days after insonification.
Fig. 8 (Coleman, Lizzi, and Jakobiec). Photo micrograph of two acute suprathreshold lesions showing adherence of insonified retina overlying one of the lesions (left). There is a small choroidal hemorrhage. The damaged scleral collagen in both lesions stains red instead of blue in this Masson trichrome preparation (X16).
Fig. 9 (Coleman, Lizzi, and Jakobiec). Photo micrograph of a healed lesion that displays thin ning of the retina, loss of photoreceptor nuclei, and placoid proliferation of pigment epithelium (hematoxylin-eosin, X100). .
VOL. 86, NO. 2
THERAPEUTIC ULTRASOUND
Fig. 10 (Coleman, Lizzi, and Jakobiec). Large lesion produced by a high energy level shows exten sive loss of sclera with fibrous replacement of episclera, fusion of choroid to scleral scar, and disor ganization of surviving retina (hematoxylin and eosin, x40).
threshold lesions is relatively spared and displays loss of its outer layers during progressive healing. Healing of the le sions is accomplished by proliferation of the retinal pigment epithelium, which fuses the remaining inner retinal layers to the fibrosed choroid (Fig. 9). In mild lesions, the sclera remains normal in thickness, but in high-energy suprathreshold lesions, the sclera becomes at tenuated or defective (Fig. 10). The defect is repaired by activation of episcleral fibroblasts, which lay down collagen. Even in massive lesions, with resultant scleral ectasia, a thin strand of surviving retina can usually be identified over the lesional zone, indicating the disparate effects of the ultrasonic energy. DISCUSSION
The clinical implications of the histopathologic findings indicate that therapeutically controlled, mild suprathreshold lesions are feasible because the ocular tunics remain intact in such lesions, without significant and lasting ultra sonic destruction of the choroid and sclera. Furthermore, the retinal tissue does not melt during the treatments, but survives in healed lesions in an inverse relationship to the energy levels used.
191
The ability of insonification to promote retinal pigment epithelium proliferation, particularly in nonpigmented cells, as demonstrated in our experiments, offers promise for treatment of hypopigmented fundi. Additionally, areas that have been treated previously but unsatisfactorily with laser or photocoagulation may be profitably retreated with insonification. These findings are even more signifi cant because these effects are possible in eyes with optically opaque media, since areas of interest can still be treated with a low-intensity ultrasound diagnostic visu alization system integrated with the treat ment system. Our results suggest: (1) it may be possi ble to produce controlled chorioretinal lesions, useful in treating retinal detach ment and other retinal degenerations; (2) the technique may be useful in dispersing vitreous hemorrhage or membranes; and (3) the production of focal necrosis of the sclera may be advantageous in the treat ment of aphakic malignant glaucoma. The potential for ultrasonic tumor treatment of lesions such as choroidal and retinal hemangioma should be explored. Treatment of malignant lesions will re quire evaluation in animal models before safety factors for the therapy of human neoplasms can be determined. Levels of insonification below the normal tissue destruction range may produce a benefi cial result. Further studies will be necessary to determine the precise energy levels re quired for the production of beneficial lesions in humans and to assure that no ill effects to surrounding tissues are concur rently produced. SUMMARY
After testing high intensity focused ul trasound on ocular tissues in animals, to obtain damage threshold equations, we used ultrasound for ocular therapy in experimental animals. We developed a therapeutic system for
192
AMERICAN JOURNAL OF OPHTHAMOLOGY
use in conjunction with a low-energy di agnostic ultrasound visualization tech nique. Our system included a spherical focused transducer operated at a frequen cy of 9.8 MHz to produce a focal spot 9 cm from the transducer surface. This technique produced controlled oc ular tissue damage similar to that pro duced by laser and xenon arc photocoagulation. It can be used to treat any level of tissue, ocular or orbital, and does not require media clarity. Specific tissue ab sorption properties necessary for laser damage were not required to produce the desired damage. REFERENCES 1. Donn, A.: Ultrasonic wave liquefaction of the vitreous humour in living rabbits. Arch. Ophthal mol. 53:215, 1955. 2. Purnell, E. W., Sokollu, A., and Holasek, E.:
AUGUST, 1978
The production of focal chorioretinitis by ultra sound. Am. J. Ophthalmol. 58:953, 1964. 3. Torchia, R., Purnell, E. W., and Sokollu, A.: Cataract production by ultrasound. Am. J. Ophthal mol. 64:305, 1967. 4. Coleman, D. J., Lizzi, F. L., Burt, W., and Wen, H.: Properties observed in cataracts produced ex perimentally with ultrasound. Am. J. Ophthalmol. 71:1284, 1971. 5. Lizzi, F. L., Burt, W., and Coleman, D. J.: Effects of ocular structures on the propagation of ultrasound in the eye. Arch. Ophthalmol. 84:635, 1970. 6. Lizzi, F . L., Coleman, D. J., Driller, J., Franzen, L.A., and Jakobiec, F. A.: Experimental, ultrasonically induced lesions in the retina, choroid, and sclera. Invest. Ophthalmol. In press. 7. Ham, W., Williams, R., Mueller, H., Guerry, D., Clark, A., and Geeraets, W.: Effects of laser radiation on the mammalian eye. Trans. NY Acad. Sci. 28:517, 1966. 8. Coleman, D. J., Konig, W. F., and Katz, L.: A hand-operated ultrasound scan system for ophthal mic evaluation. Am. J. Ophthalmol. 68:256, 1969. 9. Lizzi, F. L., Packer, A. J., and Coleman, D. J.: Experimental cataract production by high frequency ultrasound. Ann. Ophthalmol. In press.
New York, New York
The therapeutic application of con trolled ocular damage has gained wide spread use in the forms of xenon arc photocoagulation, laser photocoagulation, diathermy, and cryopexy. Each of these therapeutic modalities has specific features that make it more applicable to some situations than to others. All of these methods require precise control to produce the desired therapeutic response. A separate modality, therapeutic ultra sound, produces a similar type of tissue damage. Until recently, however, data on tissue effects and methods of control were so inadequate that such a technique could not be considered for ophthalmic treat ment. Through our laboratory studies, we have addressed these inadequacies and developed a therapeutic ultrasound sys tem for clinical use. The advantage of therapeutic ultra sound is its unique ability to produce lesions independent of the optical tissue absorption properties that limit light photocoagulation systems. This permits control of lesion placement and severity chiefly by system factors rather than by the pigmentation of the target tissue. The " b u r n " spot can be positioned in virtually any area of tissue (within the focal limits of the system), thus allowing treatment of ocular and orbital abnormalities that pre viously could be treated only after surgiFrom the Department of Ophthalmology (Drs. Coleman and Jakobiec), College of Physicians and Surgeons, Columbia University, and Riverside Re search Institute (Dr. Lizzi) New York, New York. This research was supported in part by Public Health Service Grants EY-01480 and EY-01218 from the National Eye Institute. Reprint requests to D. Jackson Coleman, M.D., E. S. Harkness Eye Institute, 635 West 165th St., New York, NY 10032.
cal exposure. The ultrasonic system af fords two advantages: it may be applied externally, as is photocoagulation; and, as is the case with radiotherapy techniques, it does not require direct visualization for the treatment of lesions. For example, patients with retinitis proliferans ob scured by opaque media or tumors ob scured by hemorrhage are candidates for ultrasonic treatment. Experience in the production of ocular lesions was described by Donn 1 ; Purnell, Sokollu, and Holasek 2 ; Torchia, Purnell, and Sokollu 3 ; Coleman and associates 4 ; and Lizzi, Burt, and Coleman. 5 These investigators attemped to produce ocular lesions chiefly in the vitreous, lens, retina, and choroid and their primary aim was to identify the minimal exposure levels re quired to produce clinically useful le sions while not damaging the adjacent ocular tissues. Difficulties in determining the proper productive energy levels and in precisely positioning the therapeutic beam had previously prevented clinical utilization of ultrasound for ocular therapy. We have studied the biologic effects of ultrasound in ocular tissues for the past five years. Meticulous control of energies required for production of corneal, lentic ular, and chorioretinal lesions has yielded threshold data on experimental animals. 6 A graph summarizing these results, as well as those salient findings of other laboratories, is shown (Fig. 1). The ener gy levels required to produce a clinically useful therapeutic lesion generally exceed energy levels of 50 W/cm 2 ; conversely, if energy levels are kept below this level in surrounding tissues, then adjacent tissue damage can be avoided. The mechanisms that cause tissue dam-
AMERICAN JOURNAL OF OPHTHALMOLOGY 86:185-192, 1978
185
186
AMERICAN JOURNAL O F OPHTHAMOLOGY ULTRASONIC INSONIFICATION OF OCULAR TISSUES o - ZISKIN
ET AL
b - BAUM c - PURNELL AND SOKOlLU d - L I 2 Z I AND COLEMAN
CATARACTSI3 5 M H z , c )
NO EFFECT 11MHz, b)
NO EFFECT 19.5 MHz, a 1
-X-
Fig. 1 (Coleman, Lizzi, and Jakobiec). Available data are superimposed to indicate ultrasonic thresh old levels for damage found in ocular tissues. The dotted line indicates the energy levels below which no tissue damage has been noted and above which the damage thresholds appear to fall.
A U G U S T , 1978
energy. The choroidal blanching proba bly reflects an alteration of blood flow that may cause the thermal reaction by reduction of heat removal. The curves for ultrasonically induced lesions compared with data derived by Ham 7 for the production of chorioretinal lesions by use of laser and xenon arc irradiation are shown (Fig. 2). The opti cally derived data used a focal spot of 0.8 mm; this is comparable to the 0.4 mmbeam we used. The similarly shaped curves suggest shared damage mecha nisms. Optical energy is absorbed chiefly in the pigment epithelium of the retina, whereas ultrasonic energy is absorbed more diffusely across the retina and choroid. Since the optic absorption coeffi cients of the pigment epithelium are extremly high as compared to that of the average ultrasonic coefficient, it is not surprising that larger amounts of ultra sonic energy are needed for similar lesion
2000
age are not completely understood, but present data suggest several phenomena may be involved in lesion formation. In chorioretinal lesions, a transitory choroi dal blanching that is easily seen in "subthreshold" lesions is consistent with a radiation pressure mechanism first sug gested by Purnell, Sokollu, and Holasek. 2 The blanching is probably produced by streaming forces that arise from absorptive gradients and radiation forces within the vitreous pathway preceding the lesion. The second form of lesion, a "threshold" lesion, is characterized by choroidal blanching, as well as an overlying edema of the retina, which ensues within 24 hours after exposure to insonification; these lesions exhibit small permanent pigment epithelial and chorioretinal ef fects. The permanent damage lesion is considered to be caused by thermal phe nomena arising from absorbed ultrasonic
1000
&
WHITE LIGHT ( \ < 9 5 0 m M )
•
ULTRASOUND ( 9 . 6 MHz)
TOO
i I
NOTE: LIGHT INTENSITY IS MULTIPLIED BY 25
400
200
100
:
^ 9 . 8 MHz ULTRASOUND (0.4mm »pol/»il«)
/
TO
- - -
^ W H I T E LIGHT (0.8 mm i p o t / i i i e )
40
20
10
, EXPOSURE MINIMAL
DURATION
(tec)
FUNDUSCOPICALLY VISIBLE CHORIORETINAL LESIONS
Fig. 2 (Coleman, Lizzi, and Jakobiec). The dam age threshold curve found with ultrasound is com pared with a damage threshold curve for chorioreti nal lesions produced by white light. The optic absorption coefficient of the pigment epithelium (white light damage) is extremely high compared to that of the average ultrasonic coefficient. Allowing for these differences in absorption thresholds, for light as compared to ultrasound, the curves are remarkably similar. The similarity of the shapes of these curves suggest a similar damage mechanism for both light and ultrasonic energy.
VOL. 86, NO. 2
THERAPEUTIC ULTRASOUND
formation. Consequently, the intensity levels for ultrasonic energy must be mul tiplied by a factor of 25 to obtain a com parable curve, but the time durations have not been scaled. With the understanding of the energy re quired to produce toxic ultrasonic effects and an understanding of the transducer design factors required to produce both the desired therapeutic effect while spar ing adjacent tissue from damage, a thera peutic system can be considered for oph thalmic application. We have developed a therapeutic sys tem intended for use in conjunction with a previously described diagnostic ultra sound visualization technique. 8 The diag nostic system uses low-energy pidsed ul trasound that produces no known tissue damage. The intensity levels of the diag nostic system are smaller than those of the therapeutic system by a factor of 10 5 , and therefore do not influence the intend ed therapeutic effects. The position of the focal zone of the therapeutic transducer appears on the diagnostic B-scan or Ascan display, or both, as an intensified area that allows manual positioning of the therapeutic focus within the targeted zone. Once the desired position is ob tained by observing the diagnostic dis play, the burn is then produced in milli second intervals in much the same manner that a laser treatment is applied. Extensive evaluation of ultrasonic lesions produced in experimental animals indi cates a unique potential for clinical appli cation in various disease states.
187
Fig. 3 (Coleman, Lizzi, and Jakobiee). The thera peutic transducer is shown with a central aperture cut out for placement of a paraxial diagnostic trans ducer. The therapeutic transducer is a focused ce ramic with a focal length of 9.0 cm and operates at a frequency of 9.8 MHz.
element (Fig. 3). A schlieren photograph of the focal zone of the therapeutic trans ducer is shown (Fig. 4). The energy levels used produce an ocu lar burn only in the focal zone. Levels ranging from 500 to 3,000 W/cm 2 in this focal zone have been used to produce lesions of the type reported here. Sur rounding regions of lower ultrasonic en ergy intensity have been found to be safe
METHOD
We designed a spherical focused trans ducer operated at a frequency of 9.8 MHz to produce a focal spot size of 0.4 mm in diameter and 4 mm in length at a distance of 9.0 cm from the transducer surface. We removed the center of the transducer to allow a diagnostic transducer to be placed paraxial with the surrounding therapeutic
Fig. 4 (Coleman, Lizzi, and Jakobiee). Schlierien photography of the therapeutic transducer beam used to produce the lesions shown in Figures 6 through 10. The focal zone is sharply angled to provide a burn of approximately 0.4 mm in diameter and 4 mm in length. The energy in surrounding areas of tissue is weak enough to produce no known damage.
188
AMERICAN JOURNAL OF OPHTHAMOLOGY
in extensive experimental studies. 9 We have demonstrated the manner in which the diagnostic and therapeutic systems are combined (Fig. 5). Both systems are manually operated, so that positioning of the beam and application of the treatment are under the physician's control. B-scan display is made by sweeping the diagnos tic transducer across the eye and the image is stored on the oscilloscope. The intensified area representing the focal zone of the therapeutic transducer is then positioned until it corresponds to the area of disease to be treated. At this point, the stored image is erased and direct kinetic diagnostic scanning is used to allow the final localization of the area of disease. In this way, the location of the therapeutic beam can be continuously observed. In clear media, vitreal, retinal, and choroidal lesions can be seen as they form. The appearance of a white area of retina visu ally resembles the lesion produced by laser or xenon photocoagulation. This system has been tested in a series of 80 albino and Dutch pigmented rab bits. The therapeutic transducer has been used to produce lesions of the cornea, lens, and the choroid, retina, sclera com plex.
AUGUST, 1978 RESULTS
Chorioretinal lesions have been an area of chief interest because the use of ultra sound in the production of such lesions is potentially of great clinical value, and the lesions can be compared with lesions produced by other modalities. Exper iments have been performed to evaluate the dimensions and nature of lesions that are created at specific ultrasonic expo sures (Figs. 6 and 7). Histologic examination of acute suprathreshold lesions demonstrated damage to the outer retina, disruption and necro sis of the pigment epithelium, subretinal exudate, mild choroidal hemorrhage, col lections of polymorphonuclear leukocytes in and around the choroidal vessels, and loss of subadjacent scleral nuclei. The trichrome stain revealed that the sclera was also altered in its staining properties, thus suggesting a physical modification of the collagen (Fig. 8). In contrast to laser-treated eyes, where most of the dam age is restricted to the level of the pig ment epithelial cells or other pigmented tissue, all layers of the ultrasonically treated eyes display some damage. The damage, however, is variable for the dif ferent tunics. The retina in mild supra-
Fig. 5 (Coleman, Lizzi, and Jakobiec). The combined therapeuticdiagnostic system as designed for clinical evaluation. The diagnostic system provides the azimuthal and axial localization, regardless of me dia clarity for positioning the treat ment zone of the therapeutic trans ducer.
Fig. 6 (Coleman, Lizzi, and Jakobiec). Fundus photograph of an albino rabbit showing a suprathreshold lesion 24 hours after insonification.
Fig. 7 (Coleman, Lizzi, and Jakobiec). Fundus photograph of an albino rabbit showing a high intensity lesion seven days after insonification.
Fig. 8 (Coleman, Lizzi, and Jakobiec). Photo micrograph of two acute suprathreshold lesions showing adherence of insonified retina overlying one of the lesions (left). There is a small choroidal hemorrhage. The damaged scleral collagen in both lesions stains red instead of blue in this Masson trichrome preparation (X16).
Fig. 9 (Coleman, Lizzi, and Jakobiec). Photo micrograph of a healed lesion that displays thin ning of the retina, loss of photoreceptor nuclei, and placoid proliferation of pigment epithelium (hematoxylin-eosin, X100). .
VOL. 86, NO. 2
THERAPEUTIC ULTRASOUND
Fig. 10 (Coleman, Lizzi, and Jakobiec). Large lesion produced by a high energy level shows exten sive loss of sclera with fibrous replacement of episclera, fusion of choroid to scleral scar, and disor ganization of surviving retina (hematoxylin and eosin, x40).
threshold lesions is relatively spared and displays loss of its outer layers during progressive healing. Healing of the le sions is accomplished by proliferation of the retinal pigment epithelium, which fuses the remaining inner retinal layers to the fibrosed choroid (Fig. 9). In mild lesions, the sclera remains normal in thickness, but in high-energy suprathreshold lesions, the sclera becomes at tenuated or defective (Fig. 10). The defect is repaired by activation of episcleral fibroblasts, which lay down collagen. Even in massive lesions, with resultant scleral ectasia, a thin strand of surviving retina can usually be identified over the lesional zone, indicating the disparate effects of the ultrasonic energy. DISCUSSION
The clinical implications of the histopathologic findings indicate that therapeutically controlled, mild suprathreshold lesions are feasible because the ocular tunics remain intact in such lesions, without significant and lasting ultra sonic destruction of the choroid and sclera. Furthermore, the retinal tissue does not melt during the treatments, but survives in healed lesions in an inverse relationship to the energy levels used.
191
The ability of insonification to promote retinal pigment epithelium proliferation, particularly in nonpigmented cells, as demonstrated in our experiments, offers promise for treatment of hypopigmented fundi. Additionally, areas that have been treated previously but unsatisfactorily with laser or photocoagulation may be profitably retreated with insonification. These findings are even more signifi cant because these effects are possible in eyes with optically opaque media, since areas of interest can still be treated with a low-intensity ultrasound diagnostic visu alization system integrated with the treat ment system. Our results suggest: (1) it may be possi ble to produce controlled chorioretinal lesions, useful in treating retinal detach ment and other retinal degenerations; (2) the technique may be useful in dispersing vitreous hemorrhage or membranes; and (3) the production of focal necrosis of the sclera may be advantageous in the treat ment of aphakic malignant glaucoma. The potential for ultrasonic tumor treatment of lesions such as choroidal and retinal hemangioma should be explored. Treatment of malignant lesions will re quire evaluation in animal models before safety factors for the therapy of human neoplasms can be determined. Levels of insonification below the normal tissue destruction range may produce a benefi cial result. Further studies will be necessary to determine the precise energy levels re quired for the production of beneficial lesions in humans and to assure that no ill effects to surrounding tissues are concur rently produced. SUMMARY
After testing high intensity focused ul trasound on ocular tissues in animals, to obtain damage threshold equations, we used ultrasound for ocular therapy in experimental animals. We developed a therapeutic system for
192
AMERICAN JOURNAL OF OPHTHAMOLOGY
use in conjunction with a low-energy di agnostic ultrasound visualization tech nique. Our system included a spherical focused transducer operated at a frequen cy of 9.8 MHz to produce a focal spot 9 cm from the transducer surface. This technique produced controlled oc ular tissue damage similar to that pro duced by laser and xenon arc photocoagulation. It can be used to treat any level of tissue, ocular or orbital, and does not require media clarity. Specific tissue ab sorption properties necessary for laser damage were not required to produce the desired damage. REFERENCES 1. Donn, A.: Ultrasonic wave liquefaction of the vitreous humour in living rabbits. Arch. Ophthal mol. 53:215, 1955. 2. Purnell, E. W., Sokollu, A., and Holasek, E.:
AUGUST, 1978
The production of focal chorioretinitis by ultra sound. Am. J. Ophthalmol. 58:953, 1964. 3. Torchia, R., Purnell, E. W., and Sokollu, A.: Cataract production by ultrasound. Am. J. Ophthal mol. 64:305, 1967. 4. Coleman, D. J., Lizzi, F. L., Burt, W., and Wen, H.: Properties observed in cataracts produced ex perimentally with ultrasound. Am. J. Ophthalmol. 71:1284, 1971. 5. Lizzi, F. L., Burt, W., and Coleman, D. J.: Effects of ocular structures on the propagation of ultrasound in the eye. Arch. Ophthalmol. 84:635, 1970. 6. Lizzi, F . L., Coleman, D. J., Driller, J., Franzen, L.A., and Jakobiec, F. A.: Experimental, ultrasonically induced lesions in the retina, choroid, and sclera. Invest. Ophthalmol. In press. 7. Ham, W., Williams, R., Mueller, H., Guerry, D., Clark, A., and Geeraets, W.: Effects of laser radiation on the mammalian eye. Trans. NY Acad. Sci. 28:517, 1966. 8. Coleman, D. J., Konig, W. F., and Katz, L.: A hand-operated ultrasound scan system for ophthal mic evaluation. Am. J. Ophthalmol. 68:256, 1969. 9. Lizzi, F. L., Packer, A. J., and Coleman, D. J.: Experimental cataract production by high frequency ultrasound. Ann. Ophthalmol. In press.
