- Email: [email protected]
ROBOTIC REMOTE LAPAROSCOPIC NEPHRECTOMY AND ADRENALECTOMY: : THE INITIAL EXPERIENCE
0022-5347/00/1646-2082/0 THE JOURNAL OF UROLOGY® Copyright © 2000 by AMERICAN UROLOGICAL ASSOCIATION, INC.®
Vol. 164, 2082–2085, December 2000 Printed in U.S.A.
ROBOTIC REMOTE LAPAROSCOPIC NEPHRECTOMY AND ADRENALECTOMY: THE INITIAL EXPERIENCE INDERBIR S. GILL,* GYUNG TAK SUNG, THOMAS H. S. HSU
AND
ANOOP M. MERANEY
From the Section of Laparoscopic and Minimally Invasive Surgery, Urological Institute and Minimally Invasive Surgery Center, Cleveland Clinic Foundation, Cleveland, Ohio
ABSTRACT
Purpose: We evaluated the feasibility of performing laparoscopic nephrectomy and adrenalectomy exclusively by using robotic telepresent technology from a remote workstation and compared outcomes with those of conventional laparoscopy in an acute porcine model. Materials and Methods: Five pigs underwent bilateral laparoscopic nephrectomy (robotic in 5 and conventional in 4) and adrenalectomy (robotic in 4 and conventional in 3). In the 9 robotic laparoscopic procedures all intraoperative manipulations were completely performed telerobotically from a remote workstation without any conventional laparoscopic assistance on site. Animals were sacrificed acutely. Results: Robotic laparoscopic nephrectomy required significantly longer total operative (85.2 versus 38.5 minutes, p ⫽ 0.0009) and actual surgical (73.4 versus 27.5 minutes, p ⫽ 0.0002) time than conventional laparoscopy. However, blood loss and adequacy of surgical dissection were comparable in the 2 groups. Robotic laparoscopic adrenalectomy required longer total operative (51 versus 32.3 minutes, p ⫽ 0.13) and actual surgical (38.5 versus 18.7 minutes, p ⫽ 0.14) time than conventional laparoscopy. The solitary complication in this study was an inferior vena caval tear during robotic right adrenalectomy, which was adequately repaired by sutures telerobotically in a remote manner. Conclusions: To our knowledge we present the initial experience with remote telerobotic laparoscopic nephrectomy and adrenalectomy. Telepresent laparoscopic surgery is feasible. KEY WORDS: kidney, nephrectomy, swine, laparoscopy, robotics
Robotic assistance during laparoscopic surgery has primarily involved remote manipulation of the laparoscope by a telementor surgeon while the actual surgical procedure is performed by a surgeon on site.1 The capability of the remote surgeon to perform operative manipulations and technical maneuvers is a recent development. Robotic assisted remote open surgery was initially performed by Bowersox and Cornum in 1996 using a surgical telemanipulator system.2 Robotic assisted remote laparoscopic surgery was initially reported by us in 1999 using the Zeus† robotic surgical system.3 In each study manual operative assistance on site was an essential component of the robotic assisted procedure. To our knowledge we report the initial experience with completely robotic laparoscopic nephrectomy and adrenalectomy performed in a remote telepresent manner. Outcome data are compared to those of conventional laparoscopic nephrectomy and adrenalectomy.
Zeus robotic arms for laparoscopic instruments were secured to the operating table (fig. 1). The surgeon then retired to a geographically separate operating room, where the robotic surgeon console was situated. No direct visualization was feasible between the 2 operating rooms (fig. 2). The robotic controller and robot connected by a hard wired connection were 15 feet apart. An assistant remained on site at the animal operating table only to exchange various robotic instruments, including the electrosurgical J-hook, spatula, tissue dissector and scissors, at the voice command of the remote surgeon and troubleshoot the robotic arms.
MATERIALS AND METHODS
Five female farm pigs weighing 30 to 40 kg. underwent bilateral nephrectomy and adrenalectomy. Selection of the robotic or conventional laparoscopic approach was done in a prospectively randomized manner. Robotic laparoscopic nephrectomy was performed on 3 right and 2 left kidneys, and conventional nephrectomy was done on 2 right and 2 left kidneys. During robotic nephrectomy a 3 port transperitoneal approach was used. A 10 to 12-inch length of 2-zero silk thread was inserted into the abdominal cavity through a port. A voice activated AESOP arm for the laparoscope and 2 Accepted for publication June 29, 2000. * Requests for reprints: Section of Laparoscopic and Minimally FIG. 1. Port placement for robotic laparoscopic nephrectomy and Invasive Surgery, Urological Institute, Cleveland Clinic Foundation, adrenalectomy with positioning of 3 robotic arms. Laparoscope (10 9500 Euclid Ave., Cleveland, Ohio 44195. mm.) is attached to AESOP arm at center, while left and right robotic † Computer Motion, Goleta, California. arms control laparoscopic instruments. 2082
ROBOTIC REMOTE LAPAROSCOPIC NEPHRECTOMY AND ADRENALECTOMY
2083
of nephrectomy or adrenalectomy. After sacrifice the excised organs were retrieved for inspection. Postoperative parameters evaluated included intra-abdominal organ injury, and inspection of the renal hilum, aorta and inferior vena cava. Statistical analysis was performed using Student’s t test for all comparisons between types of surgery except for the comparison of complications, when McNemar’s test was done. RESULTS
FIG. 2. Operating room (O.R.) setup during robotic laparoscopic surgery with surgeon located in geographically different room.
The renal artery and vein were mobilized individually. The previously inserted 2-zero silk tie was used to double ligate the renal vein and triple ligate the renal artery with 2 ties proximal and 1 distal (fig. 3). In 1 instance an aberrant renal artery was also ligated and divided. The ureter was secured and the kidney was freed circumferentially. Conventional nephrectomy was performed by standard laparoscopic technique using metallic clips for renal vessel control. Robotic laparoscopic adrenalectomy was randomly performed on 2 left and 2 right adrenal glands, and conventional adrenalectomy was done on 2 left and 1 right adrenal glands. Each adrenalectomy was performed immediately before the ipsilateral nephrectomy. The main adrenal vein was controlled by electrocautery and the adrenal gland was mobilized circumferentially. Animals were sacrificed immediately postoperatively. Intraoperative parameters evaluated included total operative time, actual surgical time, robotic setup time, blood loss, number of ligature ties placed on the renal artery and vein, complications, surgeon fatigue, and technical performance and incidents of the robot. Total operative time was defined as the time elapsed from the initial skin incision to completion of nephrectomy or adrenalectomy. Actual surgical time was defined as the time elapsed from initial laparoscopic manipulation of the intra-abdominal tissues until completion
FIG. 3. Intraoperative photograph during robotic laparoscopic nephrectomy shows 3 ligature ties placed robotically on renal artery.
Robotic laparoscopic and conventional nephrectomy was successful in all 5 and all 4 kidneys, respectively. Total surgical time was significantly longer in the robotic group (p ⫽ 0.0009). Even when robotic setup time was excluded (mean 11.8 minutes) robotic nephrectomy required significantly longer actual surgical time (p ⫽ 0.0002). This result was primarily due to several reasons. There was a learning curve. The renal vessels were suture ligated in the robotic group and clip ligated in the conventional group. Also, multiple exchanges of the robotic instruments were more timeconsuming than similar instrument exchanges during conventional laparoscopy. Blood loss was minimal (less than 5 ml.) in each group. No intraoperative complications occurred. Intraoperatively no significant malfunction of the robotic device was noted. Robotic laparoscopic and conventional adrenalectomy was successful in all 4 and all 3 adrenal glands, respectively. Although not statistically significant, total operative time and actual surgical time were almost twice as long in the robotic group. Blood loss was comparable in the groups. The only complication in this study occurred in the robotic adrenalectomy group. A small inferior vena caval tear was sustained during robotic dissection along the medial surface of a right adrenal gland. A 6 cm. 5-zero polypropylene suture was introduced into the abdomen through a 5 mm. port. The vena caval injury was remotely repaired telerobotically with intracorporeal suturing with a 100 ml. blood loss. Autopsy revealed no evidence of visceral organ injury. DISCUSSION
Sung et al recently reported the initial study demonstrating the feasibility and efficacy of robotic assisted laparoscopic pyeloplasty.3 After dissection and transection of the ureteropelvic junction by conventional laparoscopy, the pyeloplasty anastomosis was formed completely robotically. The current study is an extension of our earlier series in that all laparoscopic manipulations during nephrectomy and adrenalectomy were performed completely by remote telerobotic technology. Laparoscopic ports were inserted and instruments were attached to robotic arms manually by conventional onsite techniques. Thereafter, each intraoperative manipulation during all 9 robotic procedures (5 nephrectomies and 4 adrenalectomies), including tissue dissection, bowel mobilization and retraction, renal vessel skeletonization, ligature tie placement since robotic clip appliers are not yet available, suture cutting and electrocautery for hemostasis, were performed exclusively from a remote workstation located in a geographically different room using telerobotic technology without any manual laparoscopic assistance on site whatsoever. The only instance of on-site conventional laparoscopic manipulation in the 9 robotic procedures in this study involved the use of the suction irrigator to clear the operative field at the time of inferior vena caval injury during 1 adrenalectomy. This maneuver was necessitated by the current nonavailability of a robotic suction irrigator device. The function of the on-site assistant was exclusively to exchange various robotic end effectors (instruments) at remote surgeon request. Robotic nephrectomy took 2.7 times longer to perform than conventional laparoscopy. However, the technical adequacy of tissue dissection, renal artery and vein mobilization, blood
2084
ROBOTIC REMOTE LAPAROSCOPIC NEPHRECTOMY AND ADRENALECTOMY Intraoperative data on robotic laparoscopic versus conventional laparoscopic nephrectomy and adrenalectomy Variable
Robotic (5 specimens)
Laparoscopic nephrectomy: No. ports 3 Mean total operative time ⫾ SD (mins.) 85.2 ⫾ 2.9 Mean actual surgical time ⫾ SD (mins.) 73.4 ⫾ 3.1 Mean blood loss (cc) Less than 5 No. complications 0 Laparoscopic adrenalectomy: No. ports 3 Mean total operative time ⫾ SD (mins.) 51.0 ⫾ 1.6 Mean actual surgical time ⫾ SD (mins.) 38.5 ⫾ 1.7 Mean blood loss ⫾ SD (cc) 56.3 ⫾ 14.6 No. complications/total no. (%) 1/4 (25)* Surgeon fatigue and tremor were minimal in all subjects. * Inferior vena caval injury was repaired robotically with blood loss of approximately 100 ml.
loss and number of vascular ligature ties were satisfactory and comparable with those of conventional laparoscopy. Similarly robotic adrenalectomy took twice as long to perform as conventional laparoscopy. During robotic adrenalectomy meticulous dissection between the adrenal gland and aorta or vena cava was performed adequately. More importantly, an unanticipated, inadvertent inferior vena caval tear was repaired with sutures telerobotically from the remote workstation on an emergency basis. This capability of adequately and competently addressing an intraoperative vascular complication remotely by robotic techniques is critical if the concept of telepresence surgery is to become a clinical reality. The robotic requirements of laparoscopic pyeloplasty are somewhat different than those of nephrectomy. While the former procedure is reconstructive, the latter involves ablation of a solid organ. Thus, the degree of tissue dissection and range of instrument movement required during laparoscopic nephrectomy and adrenalectomy are much greater than those required during pyeloplasty anastomosis. Compared to our earlier study on robotic assisted pyeloplasty, during the current study laparoscopic ports were placed further apart to eliminate clashing of the robotic arms. Also, during robotic pyeloplasty increased technical precision of the movements of the instrument tips was attained by using a scale of 3.5:1. Thus, a movement of the robotic handles through a distance of 3.5 cm. resulted in a movement of the instrument tips through a distance of 1 cm. Conversely during robotic laparoscopic nephrectomy to achieve ergonomic economy of robotic handle movement while performing adequate movement of the instrument tips the scale was 1:3. Similarly during robotic adrenalectomy we used a scale of 1:2. The Zeus version 2.0 robotic surgical system used in this study comprises the 3 essential components of an ergonomically correct surgeon console, a dedicated computer controller and 3 interactive robotic arms. When the seated remote surgeon manipulates the robotic handles attached to the console, these movements are precisely filtered, scaled and relayed to the computer controller, which transmits these movements across an electromechanical interface to the robotic arms and instruments. The remote surgeon maneuvers the laparoscope by the voice controlled AESOP robotic arm. To perform the robotic procedure from a geographically different room requires the surgeon to become familiar with a complete lack of visual feedback from the operating table, a hurdle that was overcome rather rapidly in our experience. Our study identified certain limitations of the current robotic system and its end effectors (instruments). The primary shortcoming was the lack of adequate tactile feedback. For sensory immersion to be complete the tissue contact forces experienced during surgery (haptic feedback) must be seamlessly synergized with the visual feedback.2 Blunted tactile feedback compromises transparency in the surgeon-robot interface, which is a critical requirement for instinctive execution of surgical maneuvers. Automated zoom and focus function of the laparoscope as
Conventional (4 specimens)
p Value
3 38.5 ⫾ 0.5 27.5 ⫾ 1.3 Less than 5 0
1.00 0.0009 0.0002 1.00 1.00
3 32.3 ⫾ 1.5 18.7 ⫾ 1.2 30.0 ⫾ 0.0 0/3
1.00 0.13 0.14 0.10 1.00
well as 3-dimensional vision may help to facilitate surgeon visual perception with a resultant decrease in operative time. Although the electrosurgical J-hook and spatula were efficacious, currently available robotic graspers and scissors are still suboptimal. Development of a robotic suction irrigation device and a robotic clip applier is also warranted. Another factor is the cost of the robot, which is currently listed by the manufacturer at $750,000. The learning curve for robotic laparoscopic surgery is considerable. Thus, a stepwise progression of skill learning is suggested. The initial step should involve extensive practice in inanimate plastic models, which allows orientation within the remote environment, adeptness at manipulating the interactive robotic handles and an understanding of the capability of scaling robotic arm movements in a precise, graduated manner. Robotic assisted laparoscopic surgery in animal models is the next step, allowing the combination of conventional laparoscopic and robotic techniques in an animate situation. A completely robotic laparoscopic procedure represents the next level of technical difficulty. Therefore, after practicing on inanimate models we performed robotic assisted laparoscopic pyeloplasty before progressing to the completely robotic laparoscopic nephrectomy and adrenalectomy. Clinical robotic laparoscopic surgery, the final frontier, currently awaits Food and Drug Administration approval. Although proof of concept for completely robotic, telepresent laparoscopic nephrectomy and adrenalectomy has been demonstrated by this pilot study, many issues must be resolved before telepresent surgery is ready for clinical application. Time delay involving the interval from surgeon execution of an operative maneuver and its electronic relay to and from the patient remains a primary concern.4, 5 Eliminating time delay by instantaneous transmission requires widespread availability of high bandwidth communication channels. Progressing from the hard wired connection between the robotic controller and robot, as in this study, to a true telerobotic system using commercially available communication lines would be a challenging advancement. Incorporating novel software for predicting and deciphering changes in tissue mechanical properties in real time as the surgeon cuts may also help to compensate partially for the time delay.6 Surgical robotic systems incorporate human mentoring in the control loop in master-slave fashion. By augmenting human surgical sensorimotor skills with the computerized capabilities of a robot, robotic technology provides enabled remote manipulation with sensory, tactile and force feedback.7 Such computer mediation and enhancement create the potential of extending surgeon reach and dexterity beyond what is achievable by human skill only. By providing the realism of operating at the remote site movement skill is transported, resulting in distance independence.6 Such dissolution of distance and time would allow the remote surgeon not only to consult and mentor, but also perform.8 Potential users of
ROBOTIC REMOTE LAPAROSCOPIC NEPHRECTOMY AND ADRENALECTOMY
robotic technology include highly specialized physicians at tertiary care centers who may in the future extend telementoring and telesurgery facilities to select centers at geographically distant locations.
3. 4.
CONCLUSIONS
Remote telerobotic laparoscopic nephrectomy and adrenalectomy are feasible. Given anticipated advancements in telecommunication and robotic technology, clinical telepresence surgery looms on the horizon. REFERENCES
1. Janetschek, G., Bartsch, G. and Kavoussi, L. R.: Transcontinental interactive laparoscopic telesurgery between the United States and Europe. J Urol, 160: 1413, 1998 2. Bowersox, J. C. and Cornum, R. L.: Remote operative urology
5. 6.
7. 8.
2085
using a surgical telemanipulator system: preliminary observations. Urology, 52: 17, 1998 Sung, G. T., Gill, I. S. and Hsu, T. H.: Robotic-assisted laparoscopic pyeloplasty; a pilot study. Urology, 53: 1099, 1999 Funda, J., Lindsay, T. S. and Paul, R. P.: Teleprogramming: toward delay-invariant remote manipulation. Presence, 1: 29, 1992 Hu, J., Ren, J. and Sheridan, T. B.: Telerobotic surgery: stable force feedback with time delay. SPIE Proc, 2901: 2901, 1996 Thompson, J. M., Ottensmeyer, M. P. and Sheridan, T. B.: Human factors in telesurgery: effects of time delay and asynchrony in video and control feedback with local manipulative assistance. Telemed J, 5: 129, 1999 Johnston, R. S., Dietlin, L. F. and Berry, C. A.: Biomedical Results of Apollo. NASA SP-368. Washington, D. C.: United States Government Printing Office, 1975 Satava, R. M.: Robotics, telepresence and virtual reality: a critical analysis of the future of surgery. Min Inv Ther, 1: 357, 1992
Vol. 164, 2082–2085, December 2000 Printed in U.S.A.
ROBOTIC REMOTE LAPAROSCOPIC NEPHRECTOMY AND ADRENALECTOMY: THE INITIAL EXPERIENCE INDERBIR S. GILL,* GYUNG TAK SUNG, THOMAS H. S. HSU
AND
ANOOP M. MERANEY
From the Section of Laparoscopic and Minimally Invasive Surgery, Urological Institute and Minimally Invasive Surgery Center, Cleveland Clinic Foundation, Cleveland, Ohio
ABSTRACT
Purpose: We evaluated the feasibility of performing laparoscopic nephrectomy and adrenalectomy exclusively by using robotic telepresent technology from a remote workstation and compared outcomes with those of conventional laparoscopy in an acute porcine model. Materials and Methods: Five pigs underwent bilateral laparoscopic nephrectomy (robotic in 5 and conventional in 4) and adrenalectomy (robotic in 4 and conventional in 3). In the 9 robotic laparoscopic procedures all intraoperative manipulations were completely performed telerobotically from a remote workstation without any conventional laparoscopic assistance on site. Animals were sacrificed acutely. Results: Robotic laparoscopic nephrectomy required significantly longer total operative (85.2 versus 38.5 minutes, p ⫽ 0.0009) and actual surgical (73.4 versus 27.5 minutes, p ⫽ 0.0002) time than conventional laparoscopy. However, blood loss and adequacy of surgical dissection were comparable in the 2 groups. Robotic laparoscopic adrenalectomy required longer total operative (51 versus 32.3 minutes, p ⫽ 0.13) and actual surgical (38.5 versus 18.7 minutes, p ⫽ 0.14) time than conventional laparoscopy. The solitary complication in this study was an inferior vena caval tear during robotic right adrenalectomy, which was adequately repaired by sutures telerobotically in a remote manner. Conclusions: To our knowledge we present the initial experience with remote telerobotic laparoscopic nephrectomy and adrenalectomy. Telepresent laparoscopic surgery is feasible. KEY WORDS: kidney, nephrectomy, swine, laparoscopy, robotics
Robotic assistance during laparoscopic surgery has primarily involved remote manipulation of the laparoscope by a telementor surgeon while the actual surgical procedure is performed by a surgeon on site.1 The capability of the remote surgeon to perform operative manipulations and technical maneuvers is a recent development. Robotic assisted remote open surgery was initially performed by Bowersox and Cornum in 1996 using a surgical telemanipulator system.2 Robotic assisted remote laparoscopic surgery was initially reported by us in 1999 using the Zeus† robotic surgical system.3 In each study manual operative assistance on site was an essential component of the robotic assisted procedure. To our knowledge we report the initial experience with completely robotic laparoscopic nephrectomy and adrenalectomy performed in a remote telepresent manner. Outcome data are compared to those of conventional laparoscopic nephrectomy and adrenalectomy.
Zeus robotic arms for laparoscopic instruments were secured to the operating table (fig. 1). The surgeon then retired to a geographically separate operating room, where the robotic surgeon console was situated. No direct visualization was feasible between the 2 operating rooms (fig. 2). The robotic controller and robot connected by a hard wired connection were 15 feet apart. An assistant remained on site at the animal operating table only to exchange various robotic instruments, including the electrosurgical J-hook, spatula, tissue dissector and scissors, at the voice command of the remote surgeon and troubleshoot the robotic arms.
MATERIALS AND METHODS
Five female farm pigs weighing 30 to 40 kg. underwent bilateral nephrectomy and adrenalectomy. Selection of the robotic or conventional laparoscopic approach was done in a prospectively randomized manner. Robotic laparoscopic nephrectomy was performed on 3 right and 2 left kidneys, and conventional nephrectomy was done on 2 right and 2 left kidneys. During robotic nephrectomy a 3 port transperitoneal approach was used. A 10 to 12-inch length of 2-zero silk thread was inserted into the abdominal cavity through a port. A voice activated AESOP arm for the laparoscope and 2 Accepted for publication June 29, 2000. * Requests for reprints: Section of Laparoscopic and Minimally FIG. 1. Port placement for robotic laparoscopic nephrectomy and Invasive Surgery, Urological Institute, Cleveland Clinic Foundation, adrenalectomy with positioning of 3 robotic arms. Laparoscope (10 9500 Euclid Ave., Cleveland, Ohio 44195. mm.) is attached to AESOP arm at center, while left and right robotic † Computer Motion, Goleta, California. arms control laparoscopic instruments. 2082
ROBOTIC REMOTE LAPAROSCOPIC NEPHRECTOMY AND ADRENALECTOMY
2083
of nephrectomy or adrenalectomy. After sacrifice the excised organs were retrieved for inspection. Postoperative parameters evaluated included intra-abdominal organ injury, and inspection of the renal hilum, aorta and inferior vena cava. Statistical analysis was performed using Student’s t test for all comparisons between types of surgery except for the comparison of complications, when McNemar’s test was done. RESULTS
FIG. 2. Operating room (O.R.) setup during robotic laparoscopic surgery with surgeon located in geographically different room.
The renal artery and vein were mobilized individually. The previously inserted 2-zero silk tie was used to double ligate the renal vein and triple ligate the renal artery with 2 ties proximal and 1 distal (fig. 3). In 1 instance an aberrant renal artery was also ligated and divided. The ureter was secured and the kidney was freed circumferentially. Conventional nephrectomy was performed by standard laparoscopic technique using metallic clips for renal vessel control. Robotic laparoscopic adrenalectomy was randomly performed on 2 left and 2 right adrenal glands, and conventional adrenalectomy was done on 2 left and 1 right adrenal glands. Each adrenalectomy was performed immediately before the ipsilateral nephrectomy. The main adrenal vein was controlled by electrocautery and the adrenal gland was mobilized circumferentially. Animals were sacrificed immediately postoperatively. Intraoperative parameters evaluated included total operative time, actual surgical time, robotic setup time, blood loss, number of ligature ties placed on the renal artery and vein, complications, surgeon fatigue, and technical performance and incidents of the robot. Total operative time was defined as the time elapsed from the initial skin incision to completion of nephrectomy or adrenalectomy. Actual surgical time was defined as the time elapsed from initial laparoscopic manipulation of the intra-abdominal tissues until completion
FIG. 3. Intraoperative photograph during robotic laparoscopic nephrectomy shows 3 ligature ties placed robotically on renal artery.
Robotic laparoscopic and conventional nephrectomy was successful in all 5 and all 4 kidneys, respectively. Total surgical time was significantly longer in the robotic group (p ⫽ 0.0009). Even when robotic setup time was excluded (mean 11.8 minutes) robotic nephrectomy required significantly longer actual surgical time (p ⫽ 0.0002). This result was primarily due to several reasons. There was a learning curve. The renal vessels were suture ligated in the robotic group and clip ligated in the conventional group. Also, multiple exchanges of the robotic instruments were more timeconsuming than similar instrument exchanges during conventional laparoscopy. Blood loss was minimal (less than 5 ml.) in each group. No intraoperative complications occurred. Intraoperatively no significant malfunction of the robotic device was noted. Robotic laparoscopic and conventional adrenalectomy was successful in all 4 and all 3 adrenal glands, respectively. Although not statistically significant, total operative time and actual surgical time were almost twice as long in the robotic group. Blood loss was comparable in the groups. The only complication in this study occurred in the robotic adrenalectomy group. A small inferior vena caval tear was sustained during robotic dissection along the medial surface of a right adrenal gland. A 6 cm. 5-zero polypropylene suture was introduced into the abdomen through a 5 mm. port. The vena caval injury was remotely repaired telerobotically with intracorporeal suturing with a 100 ml. blood loss. Autopsy revealed no evidence of visceral organ injury. DISCUSSION
Sung et al recently reported the initial study demonstrating the feasibility and efficacy of robotic assisted laparoscopic pyeloplasty.3 After dissection and transection of the ureteropelvic junction by conventional laparoscopy, the pyeloplasty anastomosis was formed completely robotically. The current study is an extension of our earlier series in that all laparoscopic manipulations during nephrectomy and adrenalectomy were performed completely by remote telerobotic technology. Laparoscopic ports were inserted and instruments were attached to robotic arms manually by conventional onsite techniques. Thereafter, each intraoperative manipulation during all 9 robotic procedures (5 nephrectomies and 4 adrenalectomies), including tissue dissection, bowel mobilization and retraction, renal vessel skeletonization, ligature tie placement since robotic clip appliers are not yet available, suture cutting and electrocautery for hemostasis, were performed exclusively from a remote workstation located in a geographically different room using telerobotic technology without any manual laparoscopic assistance on site whatsoever. The only instance of on-site conventional laparoscopic manipulation in the 9 robotic procedures in this study involved the use of the suction irrigator to clear the operative field at the time of inferior vena caval injury during 1 adrenalectomy. This maneuver was necessitated by the current nonavailability of a robotic suction irrigator device. The function of the on-site assistant was exclusively to exchange various robotic end effectors (instruments) at remote surgeon request. Robotic nephrectomy took 2.7 times longer to perform than conventional laparoscopy. However, the technical adequacy of tissue dissection, renal artery and vein mobilization, blood
2084
ROBOTIC REMOTE LAPAROSCOPIC NEPHRECTOMY AND ADRENALECTOMY Intraoperative data on robotic laparoscopic versus conventional laparoscopic nephrectomy and adrenalectomy Variable
Robotic (5 specimens)
Laparoscopic nephrectomy: No. ports 3 Mean total operative time ⫾ SD (mins.) 85.2 ⫾ 2.9 Mean actual surgical time ⫾ SD (mins.) 73.4 ⫾ 3.1 Mean blood loss (cc) Less than 5 No. complications 0 Laparoscopic adrenalectomy: No. ports 3 Mean total operative time ⫾ SD (mins.) 51.0 ⫾ 1.6 Mean actual surgical time ⫾ SD (mins.) 38.5 ⫾ 1.7 Mean blood loss ⫾ SD (cc) 56.3 ⫾ 14.6 No. complications/total no. (%) 1/4 (25)* Surgeon fatigue and tremor were minimal in all subjects. * Inferior vena caval injury was repaired robotically with blood loss of approximately 100 ml.
loss and number of vascular ligature ties were satisfactory and comparable with those of conventional laparoscopy. Similarly robotic adrenalectomy took twice as long to perform as conventional laparoscopy. During robotic adrenalectomy meticulous dissection between the adrenal gland and aorta or vena cava was performed adequately. More importantly, an unanticipated, inadvertent inferior vena caval tear was repaired with sutures telerobotically from the remote workstation on an emergency basis. This capability of adequately and competently addressing an intraoperative vascular complication remotely by robotic techniques is critical if the concept of telepresence surgery is to become a clinical reality. The robotic requirements of laparoscopic pyeloplasty are somewhat different than those of nephrectomy. While the former procedure is reconstructive, the latter involves ablation of a solid organ. Thus, the degree of tissue dissection and range of instrument movement required during laparoscopic nephrectomy and adrenalectomy are much greater than those required during pyeloplasty anastomosis. Compared to our earlier study on robotic assisted pyeloplasty, during the current study laparoscopic ports were placed further apart to eliminate clashing of the robotic arms. Also, during robotic pyeloplasty increased technical precision of the movements of the instrument tips was attained by using a scale of 3.5:1. Thus, a movement of the robotic handles through a distance of 3.5 cm. resulted in a movement of the instrument tips through a distance of 1 cm. Conversely during robotic laparoscopic nephrectomy to achieve ergonomic economy of robotic handle movement while performing adequate movement of the instrument tips the scale was 1:3. Similarly during robotic adrenalectomy we used a scale of 1:2. The Zeus version 2.0 robotic surgical system used in this study comprises the 3 essential components of an ergonomically correct surgeon console, a dedicated computer controller and 3 interactive robotic arms. When the seated remote surgeon manipulates the robotic handles attached to the console, these movements are precisely filtered, scaled and relayed to the computer controller, which transmits these movements across an electromechanical interface to the robotic arms and instruments. The remote surgeon maneuvers the laparoscope by the voice controlled AESOP robotic arm. To perform the robotic procedure from a geographically different room requires the surgeon to become familiar with a complete lack of visual feedback from the operating table, a hurdle that was overcome rather rapidly in our experience. Our study identified certain limitations of the current robotic system and its end effectors (instruments). The primary shortcoming was the lack of adequate tactile feedback. For sensory immersion to be complete the tissue contact forces experienced during surgery (haptic feedback) must be seamlessly synergized with the visual feedback.2 Blunted tactile feedback compromises transparency in the surgeon-robot interface, which is a critical requirement for instinctive execution of surgical maneuvers. Automated zoom and focus function of the laparoscope as
Conventional (4 specimens)
p Value
3 38.5 ⫾ 0.5 27.5 ⫾ 1.3 Less than 5 0
1.00 0.0009 0.0002 1.00 1.00
3 32.3 ⫾ 1.5 18.7 ⫾ 1.2 30.0 ⫾ 0.0 0/3
1.00 0.13 0.14 0.10 1.00
well as 3-dimensional vision may help to facilitate surgeon visual perception with a resultant decrease in operative time. Although the electrosurgical J-hook and spatula were efficacious, currently available robotic graspers and scissors are still suboptimal. Development of a robotic suction irrigation device and a robotic clip applier is also warranted. Another factor is the cost of the robot, which is currently listed by the manufacturer at $750,000. The learning curve for robotic laparoscopic surgery is considerable. Thus, a stepwise progression of skill learning is suggested. The initial step should involve extensive practice in inanimate plastic models, which allows orientation within the remote environment, adeptness at manipulating the interactive robotic handles and an understanding of the capability of scaling robotic arm movements in a precise, graduated manner. Robotic assisted laparoscopic surgery in animal models is the next step, allowing the combination of conventional laparoscopic and robotic techniques in an animate situation. A completely robotic laparoscopic procedure represents the next level of technical difficulty. Therefore, after practicing on inanimate models we performed robotic assisted laparoscopic pyeloplasty before progressing to the completely robotic laparoscopic nephrectomy and adrenalectomy. Clinical robotic laparoscopic surgery, the final frontier, currently awaits Food and Drug Administration approval. Although proof of concept for completely robotic, telepresent laparoscopic nephrectomy and adrenalectomy has been demonstrated by this pilot study, many issues must be resolved before telepresent surgery is ready for clinical application. Time delay involving the interval from surgeon execution of an operative maneuver and its electronic relay to and from the patient remains a primary concern.4, 5 Eliminating time delay by instantaneous transmission requires widespread availability of high bandwidth communication channels. Progressing from the hard wired connection between the robotic controller and robot, as in this study, to a true telerobotic system using commercially available communication lines would be a challenging advancement. Incorporating novel software for predicting and deciphering changes in tissue mechanical properties in real time as the surgeon cuts may also help to compensate partially for the time delay.6 Surgical robotic systems incorporate human mentoring in the control loop in master-slave fashion. By augmenting human surgical sensorimotor skills with the computerized capabilities of a robot, robotic technology provides enabled remote manipulation with sensory, tactile and force feedback.7 Such computer mediation and enhancement create the potential of extending surgeon reach and dexterity beyond what is achievable by human skill only. By providing the realism of operating at the remote site movement skill is transported, resulting in distance independence.6 Such dissolution of distance and time would allow the remote surgeon not only to consult and mentor, but also perform.8 Potential users of
ROBOTIC REMOTE LAPAROSCOPIC NEPHRECTOMY AND ADRENALECTOMY
robotic technology include highly specialized physicians at tertiary care centers who may in the future extend telementoring and telesurgery facilities to select centers at geographically distant locations.
3. 4.
CONCLUSIONS
Remote telerobotic laparoscopic nephrectomy and adrenalectomy are feasible. Given anticipated advancements in telecommunication and robotic technology, clinical telepresence surgery looms on the horizon. REFERENCES
1. Janetschek, G., Bartsch, G. and Kavoussi, L. R.: Transcontinental interactive laparoscopic telesurgery between the United States and Europe. J Urol, 160: 1413, 1998 2. Bowersox, J. C. and Cornum, R. L.: Remote operative urology
5. 6.
7. 8.
2085
using a surgical telemanipulator system: preliminary observations. Urology, 52: 17, 1998 Sung, G. T., Gill, I. S. and Hsu, T. H.: Robotic-assisted laparoscopic pyeloplasty; a pilot study. Urology, 53: 1099, 1999 Funda, J., Lindsay, T. S. and Paul, R. P.: Teleprogramming: toward delay-invariant remote manipulation. Presence, 1: 29, 1992 Hu, J., Ren, J. and Sheridan, T. B.: Telerobotic surgery: stable force feedback with time delay. SPIE Proc, 2901: 2901, 1996 Thompson, J. M., Ottensmeyer, M. P. and Sheridan, T. B.: Human factors in telesurgery: effects of time delay and asynchrony in video and control feedback with local manipulative assistance. Telemed J, 5: 129, 1999 Johnston, R. S., Dietlin, L. F. and Berry, C. A.: Biomedical Results of Apollo. NASA SP-368. Washington, D. C.: United States Government Printing Office, 1975 Satava, R. M.: Robotics, telepresence and virtual reality: a critical analysis of the future of surgery. Min Inv Ther, 1: 357, 1992