Technique of da vinci robot-assisted anatomic radical prostatectomy

ADULT UROLOGY

TECHNIQUE OF da Vinci ROBOT-ASSISTED ANATOMIC RADICAL PROSTATECTOMY ASHUTOSH TEWARI, JAMES PEABODY, RICHARD SARLE, GURUSWAMI BALAKRISHNAN, ASHOK HEMAL, ALOK SHRIVASTAVA, AND MANI MENON

ABSTRACT Objectives. Robotic radical prostatectomy is a new procedure for treating prostate cancer. Many centers are attempting this new modality but a detailed description of the technique has not yet been published. We report the technique as performed at the Vattikuti Urology Institute. Methods. At Vattikuti Urology Institute, we have performed more than 30 such operations and have standardized the technique for safe and reproducible treatment of prostate cancer. We collected the patient data and surgical logs to improve and standardize this procedure. We recorded the operation and made relevant modifications after reviewing the recordings to improve the outcome. Results. The operation was developed on the scientific foundations of anatomic radical prostatectomy as described by Walsh and the laparoscopic prostatectomy developed at Montsouris. Our technique differs from these procedures because of the need for two surgical teams and the use of fine, endo-wrist instruments with three-dimensional stereoscopic visualization. We describe the patient setup, positioning, port placement, preparation of the robot, docking of the arms, and the surgical steps of performing anatomic prostatectomy with robotic assistance. Conclusions. This report describes the current technique of robotic prostatectomy as developed at the Vattikuti Urology Institute. UROLOGY 60: 569–572, 2002. © 2002, Elsevier Science Inc.

A

dvances in three-dimensional optics and dexterity-enhancement tools have resulted in a significant paradigm shift such that urologists no longer need to directly touch or see the structures on which they operate. Our technique of robotassisted radical prostatectomy involves the use of the da Vinci Surgical System (Intuitive Surgical, Mountain View, Calif). This is a master-slave system incorporating robotic technology and improved three-dimensional visualization, a wider range of movements, and 360° maneuverability of the tips of the instruments through the laparoscopic ports. The device offers the ability to perform surgery while sitting comfortably at a remote console, with improved ergonomics, and filtration of tremors translating into easier intracorporeal suFrom the Vattikuti Urology Institute, Henry Ford Health System, Detroit, Michigan; and Department of Urology, Case Western Reserve University School of Medicine, Cleveland, Ohio Reprint requests: Mani Menon, M.D., Vattikuti Urology Institute, Henry Ford Health System, 2F, One Ford Place, Detroit, MI 48202. Video request: Ashutosh Tewari, M.D., ash@ theEhealth.com Submitted: November 16, 2001, accepted (with revisions): May 22, 2002 © 2002, ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED

turing.1,2 Although our technique is still evolving, we have standardized the major portions of the operation, and our results are quite reproducible. This report summarizes our current approach to performing radical prostatectomy with robotic assistance, on the basis of our experience with our initial 30 patients.3 MATERIAL AND METHODS EVOLUTION Our technique is based firmly on the scientific foundations of the Montsouris technique of laparoscopic prostatectomy.4,5 However, there are important differences between the techniques of robot-assisted and conventional laparoscopic radical prostatectomy. Our modifications have been necessitated by the need for separate console-side and patient-side surgeons and subtle considerations for avoiding conflict between the da Vinci and patient-side surgeon’s ports. The ergonomics of the movements of the surgeon’s fingers have to adapt to the limitations of the currently available robotic instruments and take advantage of their versatility. Finally, our patients have a higher body mass index than European patients, and most surgeons will agree that this adds complexity to the procedure. Thus, enough differences exist between robotic and nonrobotic laparoscopic prostatectomy that a separate communication about the approach is warranted. 0090-4295/02/$22.00 PII S0090-4295(02)01852-6 569

TRAINING OF THE TEAM The robotic team (one console-side and another patientside surgeon) spent 1 month at L’Institut Mutualiste Montsouris learning the basic laparoscopic skills by working with the trainer in the animal laboratories and observing 30 consecutive cases performed by the French team. The French surgeons then trained the Detroit team in house for 30 additional cases of laparoscopic radical prostatectomy without robotic assistance. The team then trained on the da Vinci Surgical System. Various dissecting and suturing techniques were tried in vitro, in pigs, and in fresh human cadavers. A total of four fresh cadaveric laparoscopic radical prostatectomies were performed for optimal patient positioning, draping, port placement, and docking the robotic arms.

black and white balanced and calibrated to a cross bar. The initial settings are at 1:3 scaling. The height is adjusted to a comfortable level, and the vision from two eyes is merged for better viewing. The camera and arm movements are tested, and the system is left at stand by.

SETUP OF ROBOTIC ARMS (VIDEO) The robot is prepared for surgery by being draped by sterile plastic sheaths. The arms are raised and brought together before surgical docking. Additional surgical drapes are then temporarily placed on the robot to avoid inadvertent contamination.

PATIENT POSITIONING (VIDEO) PATIENT SELECTION Men with clinically localized prostate cancer who chose surgical treatment were candidates for this procedure. Patients were excluded if their life expectancy was less than 10 years or their Charlson comorbidity score was 3 or greater. Patients underwent a thorough preoperative evaluation, including serum prostate-specific antigen determination, International Prostate Symptom Score, sexual function inventory, qualityof-life score, and incontinence questionnaire. We also recorded information about other comorbidities, such as stroke, cerebral aneurysm, diabetes mellitus, hypertension, chronic obstructive pulmonary disease, and myocardial infarction. We specifically questioned patients regarding abdominal surgery, peritonitis, knee or hip surgery, and peripheral neuropathy. A history of stroke or cerebral aneurysm was a relative contraindication for this procedure, because the patient would be placed in a pronounced Trendelenburg position for 4 to 5 hours. Previous abdominal surgery was not a contraindication. Patients were admitted on the day of surgery and received deep vein thrombosis (heparin 5000 IU subcutaneously on the call to the operating room) and antibiotic prophylaxis in the preoperative holding area. Venodyne boots were placed, and the abdomen was shaved from nipple to groin.

The patient is placed in a supine position with adequate padding of the pressure points and shoulder, back, legs, and arms. We use gel pads for the patient’s back. Cotton pads are used to protect the axilla and other pressure points. We use egg crate foam padding to protect the hands while tucking them by the side of the patient’s torso. This is an important point, because the arms sometimes can come in contact with the robotic system and inadvertent injury or neurapraxia can occur. We have been able to avoid this by adequate padding and by tucking the patient’s arms while the patient is still awake. The legs are placed in the lithotomy position with enough space between the thighs for the robotic system to be wheeled in. If the patient is shorter than 6 ft., we do not use the lithotomy position and simply place the legs in a frog-leg position. The ankles are also padded. The patient’s umbilicus is cleaned with alcohol, and he is shaved from the subcostal margin to the groin. We use two egg crate forms under two criss-cross tapes to secure the patient to the table. Because of the risk of neural injury, rigid shoulder supports are not used.

PORT PLACEMENT

The Intuitive System incorporates three multi-joint robotic arms with one controlling the binocular endoscope and the other two controlling the endo-wrist instruments. Two lenses— 0° or 30°—are used. During different stages of the operation, the 30° lens can be used either looking up or down to improve the visualization. Two finger-controlled handles (the “masters”) housed in a mobile console control the two robotic arms and camera. The operating surgeon is seated at the console and does not scrub. The view of the two monitors is joined by a stereoscope, which provides excellent three-dimensional visualization. Manipulation of the masters is transmitted to a computer that filters, scales, and relays the surgeon’s movements to the robotic arms and instruments. Robotic joysticks can scale the hand movements to 1:1, 3:1, or 5:1 movement at the tip of the instruments. This scaling allows for finer and precise execution of certain steps of the operation. Tremors and small, unintended, movements as a result of holding instruments for a prolonged period are eliminated. There is no measurable delay between the movement of the handles on the console and the movement of the instruments within the patient. The instruments allow 7° of liberty in their movement, even more than the human hand.

After painting and draping the abdomen, a pneumoperitoneum is created with a Veress needle introduced through an upper left abdominal quadrant or umbilical puncture. The insufflator is set to a maximal pressure of pneumoperitoneum between 12 and 15 mm Hg. Approximately 3 to 4.5 L of gas volume is introduced before the ports are placed. A total of six ports are placed for this operation. A 12-mm port is placed at the umbilicus for the introduction of the binocular scope. The remainders of the ports are placed using the 30° “up” lens to visualize the abdominal wall. Two 8-mm ports are used for the instrument arms and are placed approximately 10 cm from the midline on a line joining the anterosuperior iliac spine to the umbilicus. Two additional ports are placed in the right side for retraction and suction purposes by the first assistant and for the insertion of sutures. The lateral one is a 10-mm and the medial one is a 5-mm port. We usually use a sixth assistant port (5 mm) on the left side. This port is placed laterally in the flank slightly inferior to the left robotic port. We then dock the camera arm to the central 12-mm port and the instrument arms to the 8-mm ports. The patient-side surgeons, standing on left and right of the patient, carefully control the port and robotic arms to achieve proper alignment and precise docking. Care is taken to avoid injury to the abdominal skin and to the legs and hands by inadvertent compression by either robotic arm. It takes approximately 15 to 45 minutes from introduction of the Veress needle to the actual beginning of surgery.

SETUP OF CONSOLE (VIDEO)

INSTRUMENTS (VIDEO)

The console is started and primed. The system is then allowed to go through the self-testing process during which it recognizes various components. The camera and lens are

Conventional laparoscopic instruments are used by the patient-side surgeon and include atraumatic graspers, scissors, suction, bipolar cautery, and intracorporeal clips. The dissec-

da Vinci ROBOTIC TECHNOLOGY (VIDEO)

570

UROLOGY 60 (4), 2002

tion is done using just two robotic instruments: a monopolar hook on the right side and forceps on the left (for a righthanded surgeon). Two needle drivers are used during anastomosis. The electrosurgical device is set to 30 to 50 power setting for coagulation.

SURGICAL STEPS (VIDEO) Lymph Node Dissection. Lymphadenectomy is performed using the 0° lens for visualization. We prefer to use a wide field of vision so that the major vessels are always in the field. A 1:3 scaling is used so that the dissection is more precise. Although an additional 15 to 20 minutes are needed for this step, we gain some time for the subsequent anterior vesical dissection. Posterior Dissection. In accordance with the Montsouris technique, retrovesical dissection of the seminal vesicles and vas deferens was the first step of surgery. We have recently stopped doing this step and instead start the dissection anteriorly. We prefer to use the 30° “down” lens for this part of the procedure. As is seen in the video clip, the parietal peritoneum covers the urinary bladder anteriorly and rectum posteriorly. Between these two structures lie the vasa deferentia and seminal vesicles. It is important to recognize the vas deferens early and to stay on its surface until it leads to the seminal vesicles posteroinferiorly. The vas deferens and seminal vesicles are usually encased in whitish areolar tissue and the detection of fat during the search for these structures usually signifies that one is in the wrong location. The deferential artery and seminal vesicle pedicles (at the tip) are controlled using either a bipolar forceps or a clip. Care is taken to avoid using excessive electrical currents, because the neurovascular bundles lie very close to the tips of the seminal vesicles.We have introduced one modification for this step of the dissection. It is sometimes possible to mistake the ureter for the vas deferens in the very narrow field of vision at this stage of the operation. To avoid inadvertent damage to the ureter, we do not transect the vas deferens until we have identified the posterior surface of the seminal vesicles adjacent to the vas deferens. Second, after dissecting anterior to the seminal vesicles, we always incise the fascia anteriorly. This helps significantly in the delivery of the seminal vesicles and vas deferens after subsequent posterior bladder neck transection. Incision and Mobilization of Bladder and Prostate—Development of Space of Retzius (Video). We use the 0° lens and a 1:1 nonscaling mode for this part of the operation. We start as high on the anterior abdominal wall as our port placement will allow and make a wide inverted U incision around the bladder (which is filled with 150 mL saline at this stage). The medial umbilical ligaments are carefully cut using a hook or scissors, and dissection usually continues to the urachus. Exposure of Prostatic Apex and Endopelvic Fascia. We use the 0° lens with a 1:3 scaling for this part of the dissection. We sweep the tissue away from the pubic symphysis, expose the endopelvic fascia and puboprostatic ligaments, and mobilize the levator fibers off the prostate to clear a space around the apex. Several venous tributaries travel in this area—sometimes called the “veins of Kelley”—and require early control with bipolar cautery. The endopelvic fascia often has a weak area—the “gap of Guy”—through which we can expose the levator fibers. We then tease the fatty tissue off the anterior surface of the prostate. The point at which this tissue does not separate easily from the prostate is the prostatovesical junction. Dorsal Vein Stitch. In our experience, the 0° or the 30° “up” lens should be used for the dorsal vein stitch without scaling. We use a laparoscopic-length suture (6 in.) on a CT-1 needle (0 braided, polyglactin suture with a 36-mm taper needle; Ethicon) to control the dorsal venous plexus with two simple stitches. UROLOGY 60 (4), 2002

Bladder Neck Transection. We find that the 30° “down” da Vinci lens aids enormously in the precise dissection of the bladder neck. The prostatovesical junction is usually at the point at which loose fat can no longer be swept off the prostate. With experience, one can identify a shallow groove between the prostate and bladder and the horizontally oriented detrusor fibers. Using an electrocautery hook, we develop this plane and dissect the bladder away from the prostate. Sometimes the prostatovesical junction is demarcated better laterally than at the midline. The bladder neck is incised to expose the Foley catheter. The balloon is deflated and pulled anteriorly to expose the posterior bladder neck, which is incised. The anterior layer of Denonvilliers fascia is exposed. Note that this has been incised at an earlier phase of the dissection. This helps in proper orientation and reduces the likelihood of either entering the prostate or undermining the bladder neck. The vasa deferentia are easily visualized through this window. The vas deferens and seminal vesicles are grasped and pulled upward. The remaining attachments between the bladder and prostate are divided with electrocautery to expose the lateral pedicles of the prostate. Lateral Pedicle Control and Nerve Sparing. Using blunt and sharp dissection, we expose the pedicles and control them using Hem-o-lok polymer ligating clips (Weck Systems, Triangle Park, NC). If nerve sparing is planned, we enter the plane between the layers of prostatic fascia. The neurovascular bundle is enclosed within layers of prostatic and periprostatic fascia. We reflect the periprostatic fascia and dissect away the neurovascular bundle.6 We find that the “wristed” instruments of the da Vinci System aid this dissection enormously and help us to dissect the nerves away from the specimen precisely. This dissection is carried as far downward as possible, and, in any event, to the lateral convexity of the prostate. Dissection Behind the Prostate. Once both the vas deferens and seminal vesicles have been dissected free, they are pulled upward by the left-side assistant. This maneuver places the Denonvilliers fascia at tension, and a faint plane between the rectum and prostate is visible. The fascia is sharply incised to expose the perirectal fat. Using sharp and blunt dissection, we develop this plane to dissect the prostate and Denonvilliers fascia off the rectum. The distal limit of this dissection is the prostatic apex. Apical Dissection of the Prostate. At this point of the operation, three structures attach the prostate: the previously ligated dorsal vein complex, the puboperinealis muscle, and the urethra and attached rectourethralis muscle.We use a 0° lens with 1:3 scaling to incise the dorsal venous complex and urethra. Using an electro-hook, the dorsal complex is incised tangential to the prostate to avoid capsular incision. A plane between the urethra and dorsal venous complex is gently developed to expose the anterior urethral wall. The puboperinealis muscle is bluntly dissected away from the urethra, taking care to push back and preserve the neurovascular bundle. A van Buren urethral sound is placed precisely at the midline to identify the anterior surface of the urethra at the urethroprostatic junction. To minimize the possibility of a positive apical margin, the anterior wall of the urethra is transected with the scissors 5 to 10 mm distal to the apex of the prostate. The posterior wall of the urethra and the rectourethralis muscle are cut. The freed specimen is then placed in an EndoCatch (Ethicon) specimen retrieval bag. Urethrovesical Anastomosis. The 0° lens is used for visualization, and no scaling is used. We use an RB1 (9-in., 2-0 braided, polyglactin suture on a 17-mm taper needle; Ethicon) stitch for the anastomosis between the bladder neck and urethra. Endo-wrist technology helps significantly in this stage of surgery, because the sutures can be placed at almost any angle. If the bladder neck has been preserved, we use interrupted sutures. If the bladder neck is wide, a running anastomosis is 571

preferred. A 20F Foley catheter is introduced and inflated to 20 to 30 mL. The bladder is filled with 250 mL saline, with the balloon of the catheter away from the bladder neck, to test the integrity of the anastomosis. Specimen Retrieval and Surgery Completion. A suction drain is placed through one of the 5-mm ports. The 10-mm port is closed and the entrapped specimen removed after enlarging the umbilical port incision as required. The incision is closed in layers.

POSTOPERATIVE CARE Patients usually ambulate and/or tolerate clear fluids on the same evening or the next morning. Solid food is only started after they have passed flatus. The suction drain is removed within 24 hours, and the patient is discharged. A gravity cystogram is performed on the fourth postoperative day, and catheter is removed if no leakage is noted.

CONCLUSIONS Robot-assisted anatomic prostatectomy is a safe, effective, and reproducible procedure for the management of clinically localized prostate cancer. We find that the classic Montsouris technique for laparoscopic radical prostatectomy can be used for this procedure, with modifications as required by robotic technology. In our hands, the procedure is in evolution and further modifications will be introduced as newer instruments—and our skills—are developed.

572

ACKNOWLEDGMENT. To Dr. Bertrand Guillonneau and Guy Vallancien for mentoring us in this procedure; and to Dr. I. S. Gill for providing constructive comments. REFERENCES 1. Abbou CC, Hoznek A, Salomon L, et al: Laparoscopic radical prostatectomy with a remote controlled robot. J Urol 165: 1964 –1966, 2001. 2. Binder J, and Kramer W: Robotically-assisted laparoscopic radical prostatectomy. BJU Int 87: 408 –410, 2001. 3. Menon M, Tewari A, Baize B, et al: Prospective comparison of radical retropubic prostatectomy and robot-assisted anatomic prostatectomy: the Vattikuti Urology Institute experience. Urology (in press). 4. Guillonneau B, and Vallancien G: Laparoscopic radical prostatectomy: the Montsouris technique. J Urol 163: 1643– 1649, 2000. 5. Guillonneau B, Cathelineau X, Doublet JD, et al: Laparoscopic radical prostatectomy: the lessons learned. J Endourol 15: 441–448, 2001. 6. Tewari A, Menon M, Peabody J, et al: An anatomic map to assist identification of the neurovascular bundle during laparoscopic radical prostatectomy: a study of twelve fresh male cadavers, in Herman AK (Ed): Contemporary Trends in Laparoscopic Urologic Surgery. New Dehli, BI Churchill Livingston, 2002, pp 297–300.

A video clip of this procedure can be viewed on the Internet at: http://www.goldjournal. net.

UROLOGY 60 (4), 2002