Focused ultrasound ablation of the epididymis with use of thermal measurements in a canine model

Transcription

1 FERTILITY AND STERILITY VOL. 78, NO. 3, SEPTEMBER 2002 Copyright 2002 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Focused ultrasound ablation of the epididymis with use of thermal measurements in a canine model Nathaniel M. Fried, Ph.D., a William W. Roberts, M.D., a Yegor D. Sinelnikov, Ph.D., c E. James Wright, M.D., a and Stephen B. Solomon, M.D. a,b Johns Hopkins Hospital, Baltimore, Maryland Received December 10, 2001; revised and accepted March 6, Supported by a research grant from Transurgical, Inc., Setauket, New York. N. Fried and S. Solomon hold stock options in Transurgical, Inc. and serve as consultants for the company. Reprint requests: Nathaniel Fried, Ph.D., Department of Urology, Johns Hopkins Bayview Medical Center, 4940 Eastern Avenue, Building A, Room 347c, Baltimore, Maryland (FAX: ; E- mail: nfried@jhmi.edu). a Department of Urology, Johns Hopkins Medical School. b Department of Radiology, Johns Hopkins Medical School. c Transurgical, Inc., Setauket, New York /02/$22.00 PII S (02) Objective: To explore the epididymis as an alternative anatomical target to the vas deferens for noninvasive male sterilization using therapeutic focused ultrasound. Design: Controlled preclinical study. Setting: Canine animal model in an academic research environment. Patient(s): Four healthy male mongrel dogs (30 35 kg). Intervention(s): A transducer mounted on a plastic clip delivered ultrasound energy to the canine epididymis. Thermocouples placed transcutaneously into the epididymis, intradermally, and on the skin surface recorded temperatures during ablation with a wide range of acoustic powers and sonication times (control, 3 W/120 s, 5 W/90 s, 7 W/60 s). Main Outcome Measure(s): Thermocouple temperature measurements determined the optimal range of ablation parameters that produced successful thermal occlusion of the epididymis without adverse effects (e.g., skin burns, testicular injury). Result(s): A large therapeutic window was determined (power 3 7 W, time seconds) over which noninvasive thermal occlusion of the epididymis can be achieved. Thermal occlusion rates were higher, and complications lower, than found previously with vas deferens ablation. Conclusion(s): The epididymis represents a larger and easier target than the vas deferens for performing noninvasive male sterilization using focused ultrasound. Long-term azoospermia studies will be necessary to confirm permanent sterilization with this technique. (Fertil Steril 2002;78: by American Society for Reproductive Medicine.) Key Words: Ultrasound, HIFU, epididymis, sterilization, vasectomy, skin burns, temperature, noninvasive Surgical sterilization is the most commonly used method of contraception among couples of reproductive age in the United States (1). Although male sterilization is less expensive, has a higher success rate, has fewer complications, and is easier to perform than female sterilization, tubal ligation remains a more commonly performed procedure than vasectomy (1 3). This discrepancy may be due to a number of factors, including societal pressures and fears of complications regarding vasectomy, which include bleeding, infection, and scrotal pain (1 6). A noninvasive, safer, and faster method of male sterilization may reduce these concerns and result in more widespread use of male sterilization. Previous studies in our laboratory have shown that therapeutic focused ultrasound may be used to successfully thermally occlude the vas deferens, potentially providing a completely noninvasive method of vasectomy (7, 8). However, inconsistent vas occlusion caused by problems with co-location of the vas within the ultrasound focus, and burns due to strong absorption of ultrasound energy in the skin, remain challenges to perfecting this technique. In this study, we apply our ultrasound device to a new and potentially simpler target in the male reproductive tract, the epididymis. There are several reasons why the epididymis may be an easier target than the vas deferens for noninvasive male sterilization using focused ultrasound. First, the epididymis is much 609

2 larger than the vas deferens: the epididymis consists of a bundle of very small ( 200- m diameter) epididymal ducts wrapped into an 2-cm diameter sphere at the epididymal head (9). Theoretically, a thermal lesion placed in the epididymis would occlude the same duct at several locations, providing multiple sites of occlusion. The vas deferens, on the contrary, is a single muscular 2-mm outer diameter tube, a much smaller target, making precise co-location of the ultrasound focus within the vas crucial to successful thermal occlusion (9). Furthermore, the larger epididymal target allows placement of the ultrasound focus deeper into the subsurface tissue, allowing a tissue buffer between the skin and target site, and thus reducing the probability of skin burns. Second, the epididymis is a more delicate structure than the vas, which is covered by a thick, muscular wall, making the epididymis more susceptible to thermal injury and occlusion than the vas. Thus, successful ablation could potentially be achieved with lower power, again reducing the probability of skin burns. In this study, experimental thermocouple temperature measurements were performed to help determine the optimal range of ablation parameters necessary for achieving thermal occlusion of the epididymis without causing skin burns. Current observations with focused ultrasound ablation of the epididymis were then compared with previous results of vas deferens ablation. METHODS Ultrasound System A 100-W radiofrequency power supply was interfaced with a laptop personal computer and operated with Labview software (National Instruments, Austin, TX). The two-radian (114.6 degrees) ultrasound transducer had dimensions of 1-cm width by 4-cm length, with a 2-cm radius, and was made of an inexpensive and disposable proprietary polymer material (Transurgical, Inc., Setauket, NY). The transducer was designed to produce 4 MHz ultrasound energy, creating a cigar-shaped FWHM intensity profile in the epididymis of approximately mm (width depth length) at the focus. The transducer was mounted into a plastic clip with movable jaws for grasping the epididymal head (Fig. 1), an approach similar to that used previously with the vas deferens. A custom-built heat exchanger was used to circulate chilled (10 2 C), degassed water through a latex balloon attached to the clip on the front surface of the transducer. An acoustic power meter (model UPM-DT-10E, Ohmic Instruments, Easton, MD) was used to calibrate the ultrasound clips in degassed water and measure the transducer acoustic output power both pre- and postablation. The surface intensity was calculated as the acoustic output power divided by the surface area of the transducer. FIGURE 1 Photograph of experimental setup demonstrating application of the ultrasound clip and thermocouples to the epididymis. Animal Preparation All animal work was performed in accordance with the guidelines of the Animal Care and Use Committee at Johns Hopkins University. Mongrel dogs weighing kg were premedicated with an intramuscular injection of acepromazine (0.75 mg/kg). An endotracheal tube was then inserted and the dog placed on respiratory support with maintenance anesthesia of 1% 2% isofluorane in oxygen. The scrotal region was then shaved with an electric razor to provide good acoustic coupling. The epididymal head was placed between the jaws of the ultrasound clip by a trained urologist (W.W.R.) with the focus centered inside the epididymis (Fig. 1). Ultrasound gel was then applied to the skin surface and the balloon manually inflated with a syringe to insure good skin contact and acoustic coupling. Insulated micro-thermocouples (Cu-Cn, 125- m diameter, 30-millisecond response time; Omega Corporation, Stamford, CT) were guided to the epididymis, intradermal (subsurface skin), and skin surface (at the skin balloon interface) through 16-gauge syringe needles. Once positioned, the needles were removed, and the thermocouples were interfaced to a personal computer with data acquisition software (Omega Corp.) for automated acquisition and realtime display of temperature vs. time data. Temperatures were measured once every 33 milliseconds with an accuracy of 2 C. Thermocouple Measurements Initial preablation cooling experiments were conducted to determine the length of time necessary to cool the skin surface to approximately the chilled water temperature (10 2 C). A preablation cooling period of 60 seconds was used during all of the ablation experiments. The first 10 seconds of each experiment recorded the baseline tissue 610 Fried et al. Ultrasound ablation of epididymis Vol. 78, No. 3, September 2002

3 temperatures, followed by 60 seconds of preablation water cooling. At t 70 seconds, sonication was initiated for varying times between 60 and 120 seconds. After sonication was completed, active postablation water cooling continued until tissue temperatures returned to approximately their baseline levels. Experiments were performed in a total of four dogs with a single set of ablation parameters (precooling control, 3 W/120 s, 5 W/90 s, or 7 W/60 s) being used for the left and right epididymis (n 2) on each dog. The ultrasound transducer surface intensities were 0.8, 1.4, and 1.9 W/cm 2, respectively, for the chosen ablation parameters. These parameters were similar to those used in previous vas deferens ablation studies, therefore direct comparisons could be made. Representative thermocouple temperature curves were plotted for each set of ablation parameters. Statistical analysis was not performed in this preliminary study due to the low n-values (n 2) for each set of ablation parameters. During the animal recovery period, the ablation sites were examined grossly for evidence of skin burns. After 21 days, the testicles and epididymis were excised and processed using standard histologic procedures and hemotoxylin and eosin stains. FIGURE 2 Representative thermocouple measurements at the epididymis, intradermal, and skin surface for intermediate ablation parameters (P 5W,t 90 seconds, SI 1.4 W/cm 2 ). Temperature data are separated into several phases: baseline (10 seconds), preablation cooling (60 seconds), sonication ( seconds), and postablation cooling. The therapeutic window is defined as the time difference between when the intradermal skin and epididymis reach a coagulation temperature of 52 C. RESULTS Thermocouple Measurements Representative temperature time data was taken using ablation parameters similar to previous vas deferens ablation experiments (control, 3 W/120 s, 5 W/90 s, 7 W/60 s). Control studies were performed in which chilled water was rapidly circulated through the balloon to help determine the optimal preablation cooling times. The temperatures at the skin balloon interface and intradermally decreased rapidly, producing a large thermal gradient between the target epididymis and intervening skin layers within seconds. After 60 seconds, there was a significant thermal gradient of 15 C between the epididymal temperature of 34 C and the intradermal skin layers at 19 C. Precooling times longer than 120 seconds provided decreasing benefit at the expense of increased operative times. Temperature time curves for epididymal ablation at low transducer acoustic powers (3 W/120 s) were measured. Peak epididymis temperatures reached 60 C, whereas intradermal temperatures remained below 40 C. The therapeutic window was very large at these parameters because the target epididymis reached and stayed above coagulation temperatures of 52 C for an extended period of time, while skin temperatures remained well below the coagulation threshold indefinitely. In Figure 2, results are shown for ablation with intermediate acoustic powers of 5 W/90 s. Peak temperatures at the epididymis reached 65 C, whereas intradermal skin temperatures continued to increase, reaching the coagulation threshold of 52 C just as sonication was stopped. These temperature trends indicate that a large therapeutic window of 60 seconds exists, corresponding to a sonication time of 90 seconds before skin burns should start to appear. Epididymal ablation performed with high acoustic powers of 7 W/60 s resulted in a decreasing therapeutic window. Both epididymal and intradermal temperatures rapidly increased, reaching 67 C and 57 C, respectively, before sonication was stopped. As a result, the therapeutic window was only about 10 seconds, corresponding to a total sonication time of 20 seconds. Gross and Histologic Observations There was no evidence of skin burns at the ablation sites during gross observation of the scrotum immediately after ablation. There were also no observable signs of pain by the dogs either during the procedure or during postoperative monitoring. Successful targeting and creation of a thermal lesion in the epididymal head was confirmed by the presence of a hard nodule beneath the skin during manual examination. A histologic section of the ablated epididymis and adjacent testicular tissue at 21 days is shown in Figure 3. A large area of the epididymis was thermally occluded, as demonstrated by the presence of a dense region of scar tissue. In addition, healthy epididymal tissue adjacent to the targeted area shows dilated ducts (presumably due to back pressure from the obstructed site), demonstrating successful occlusion. FERTILITY & STERILITY 611

4 FIGURE 3 Representative H&E-stained histologic photograph of the epididymis ablated with a power of 5 W for 60 seconds, shown here 21 days after ablation. T testicle; E epididymis; S scar tissue. Note that the epididymal ducts are dilated due to obstruction of the coagulated epididymis. The testicle appears normal. DISCUSSION Main Findings Thermocouple experiments suggest that a large therapeutic window exists at which the epididymis may be thermally occluded without injury. This wide range of ablation parameters encompasses acoustic transducer powers of 3 7 W, surface intensities of W/cm 2, and sonication times of seconds. Preablation cooling studies also suggest that cooling times between 60 and 120 seconds produce a significant thermal gradient between the skin layers and the epididymis. This thermal gradient partially offsets the effects of selective skin absorption of ultrasound energy during sonication, thus protecting the skin from burns. Comparison of Epididymis and Vas Deferens Previous studies using focused ultrasound ablation of the vas deferens resulted in a number of problems, including skin burns and inconsistent vas occlusion (7, 8). Burns were primarily due to high ultrasound absorption in the skin layers, and the low occlusion rates due to difficulty in targeting the vas and delivering sufficient thermal energy for occlusion (7). Our successful results with focused ultrasound ablation of the epididymis may be attributed primarily to significant anatomical differences between the two targets. The epididymis represents an easier target than the vas deferens for several reasons. First, the epididymal head is larger than the vas (2-cm diameter vs. 2-mm diameter), not only eliminating co-location issues, but also allowing deeper targeting and less probability of skin burns. No skin burns were observed during epididymal ablation due to our ability to grasp more tissue between the jaws of the clip, thus providing a thermal buffer between the target epididymis and the skin. The reduced skin thickness layer above the epididymis as compared to the vas ( mm vs mm) also minimized burns due to direct skin absorption. These differences in anatomy probably explain the large differences observed in the peak target temperatures during sonication. For example, at intermediate acoustic powers (5 W/90 s), peak epididymis and vas temperatures reached 65 C and 90 C, respectively (7). Second, the individual epididymal ducts are much easier to occlude than the vas. The ducts have a smaller lumen than the vas (0.2 mm vs. 0.3 mm), and are also more delicate than the vas, which is insulated by a thick, muscular wall. A lesion placed in the epididymis would potentially occlude the same duct at several different locations, insuring success, whereas problems co-locating the focus of the ultrasound transducer with the single 2-mm diameter vas deferens resulted in frequent misses of the target. These differences may help explain why occlusion rates measured approximately 83% (10/12) when targeting the epididymis, and only 40% (4/10) when targeting the vas deferens during previous studies. In addition, complication rates (skin burns) measured 40% when targeting the vas deferens and 0% when targeting the epididymis (8). Study Limitations There are several limitations to this preliminary study, which need to be addressed. First, optimization of the treatment parameters and histologic evidence supporting thermal occlusion of the epididymis represent significant advances toward development of a noninvasive vasectomy procedure. However, definitive evidence of male sterilization will ultimately need to be confirmed through ejaculation studies in animals with sperm count measurements. The optimal ablation parameters identified in this study will be used in a long-term animal study with assessment of azoospermia in ejaculate specimens to determine successful sterilization. Second, it should be emphasized that an epididymal approach to male sterilization may limit the reversibility of this procedure. Although this may appear to be a significant drawback of our approach, we believe that the availability and need of a truly reversible male sterilization procedure is overstated, for several reasons. First, only a small percentage (2% 6%) of couples actually choose to attempt a vasectomy 612 Fried et al. Ultrasound ablation of epididymis Vol. 78, No. 3, September 2002

5 reversal (9, 10). This small number is probably due to a combination of the high cost of vasectomy reversal as well as the lack of interest. Second, of the number of couples choosing vasectomy reversal, only a fraction (40% 60%) actually experience a successful vasectomy reversal as judged by pregnancy (10 12). Thus, only 1% 3% of couples would be adversely effected by a nonreversible male sterilization technique. For these reasons, we do not believe that reversibility should be a major consideration when developing a new approach to sterilization. Finally, although the canine model used in these studies is the closest convenient animal model to humans for developing a new noninvasive male sterilization technique, we recognize that significant limitations of this model exist. Although our studies support the observation that canine skin is considerably thicker than human skin on average, this difference may contribute to increased ultrasound absorption and overstate the problems with skin burns that would be encountered in humans. Furthermore, the skin thickness measurements reported in this study are only valid for direct comparison between our epididymis and vas deferens results, and cannot be generalized to humans, or even other dog types. Nevertheless, comparative results show that epididymal ablation is easier, more successful, and has fewer adverse effects than vas ablation under comparable conditions. In conclusion, this study has demonstrated that the epididymis may be a simpler target for thermal occlusion than the vas deferens when developing a noninvasive method of male sterilization using therapeutic ultrasound. A large therapeutic window was observed with acoustic powers of 3 7 W and sonication times of seconds, in which successful occlusion of the epididymis can be achieved without adverse effects (e.g., skin burns, testicular injury). Future research will use this optimal range of ablation parameters to conduct long-term, azoospermia studies in animals to confirm permanent sterilization after focused ultrasound ablation. Acknowledgments: The authors thank Patrick Lopath, Todd Fjield, Paul Harhen, and David Acker of Transurgical, Inc. (Setauket, NY), who developed the novel ultrasound clip used in these experiments. They also thank Louis Kavoussi, M.D., of the Department of Urology for his clinical advice and Theresa Chan, M.D., of the Department of Pathology for her help in analyzing the histology. References 1. Chandra A. Surgical sterilization in the United States: prevalence and statistics, Vital Health Stat 1998;23: Miller WB, Shain RN, Pasta DJ. Tubal sterilization or vasectomy: how do married couples make the choice. Fertil Steril 1991;56: Hendrix NW, Chauhan SP, Morrison JC. Sterilization and its consequences. Obstet Gynecol Surv 1999;54: Shain RN, Miller WB, Holden AE. Factors associated with married women s selection of tubal sterilization and vasectomy. Fertil Steril 1985;43: Alderman PM. Complications in a series of 1224 vasectomies. J Fam Pract 1991;33: Schwingl PJ, Guess HA. Safety and effectiveness of vasectomy. Fertil Steril 2000;73: Fried NM, Sinelnikov YD, Pant BB, Roberts WW, Solomon SB. Noninvasive vasectomy using a focused ultrasound clip: thermal measurements and simulations. IEEE Trans Biomed Eng 2001;48: Roberts WW, Chan DY, Fried NM, Wright EJ, Nicol T, Jarrett TW, et al. High intensity focused ultrasound ablation of the vas deferens in a canine model. J Urol 2002;167: Goldstein M. Surgical management of male infertility and other scrotal disorders. In: Walsh PC, Retik AB, Vaughan ED, Wein AJ, eds. Campbell s urology, Vol. 2. Philadelphia: WB Saunders, 1998: Potts JM, Pasqualotto FF, Nelson D, Thomas AJ Jr, Agarwal A. Patient characteristics associated with vasectomy reversal. J Urol 1999;161: Heidenreich A, Altmann P, Neubauer S, Engelmann UH. Microsurgical vasovasostomy in the age of modern reproduction medicine. A costbenefit analysis. Urologie A 2000;39: Holman CD, Wisniewski ZS, Semmens JB, Rouse IL, Bass AJ. Population-based outcomes after 28,246 in-hospital vasectomies and 1,902 vasovasostomies in Western Australia. BJU Int 2000;86: FERTILITY & STERILITY 613