Laparoscopy and Robotics Matched Comparison of Robotic vs Laparoscopic Nephroureterectomy: An Initial Experience Sapan N. Ambani, Alon Z. Weizer, J. Stuart Wolf, Jr, Chang He, David C. Miller, and Jeffrey S. Montgomery OBJECTIVE
METHODS
RESULTS
CONCLUSION
To compare our initial robotic-assisted nephroureterectomy (RAN) experience with a wellestablished practice of performing laparoscopic nephroureterectomy (LN) to treat upper tract urothelial carcinoma (UTUC). We reviewed our prospectively maintained minimally invasive surgery database. Patients who underwent RAN from April 2009 to December 2011 were matched by pathologic tumor stage and age (10 years) to those who underwent LN. Twenty-two matched pairs were evaluated. Mean operative time (298 vs 251 minutes) and estimated blood loss (EBL, 380 vs 233 mL) were significantly higher for RAN, with a greater need for transfusion in this group. Complication rates were similar. The RAN group trended toward more frequent lymph node dissection and greater median node count when lymph node dissection was performed (59% vs 27% [P ¼ .07] and 5.5 vs 1.0 [P ¼ .13]). After a median followup of 10 months for RAN and 15 months for LN, no significant difference was seen in the rate of bladder (36% vs 37%) or distant (32% vs 23%) recurrence, with similar median time to any recurrence (9 months vs 4 months, P ¼ .32). RAN was associated with higher operative time and blood loss likely because of more frequent use of node dissection, robot repositioning, and technical inexperience. Lymph node dissection was more frequently performed with RAN, which reflects surgeon practice patterns. When a lymph node dissection was performed, the median node count was greater with RAN. Our initial experience with RAN suggests that this is an acceptable approach for the management of UTUC. UROLOGY 83: 345e349, 2014. 2014 Elsevier Inc.
pper tract urothelial carcinoma (UTUC) is rare, accounting for only 5% of all urothelial malignancies.1 As a result, the nuances of the appropriate management of this disease are not fully established. The historic standard management of UTUC is an open nephroureterectomy with excision of a bladder cuff to remove the entire urothelium of the affected renal unit. More recently, there has been an emphasis on using laparoscopic techniques to perform this procedure.2 Oncologic equivalence and improvement in postoperative convalescence has been shown when comparing laparoscopic nephroureterectomy (LN) with open nephroureterectomy,2-6 indicating that LN is an
U
effective treatment option. The advent of robotic technology has expanded the potential for the minimally invasive management of urologic conditions. Laparoscopic robotic-assisted nephroureterectomy (RAN) for UTUC was initially performed in combination with traditional laparoscopy.7,8 Eandi et al9 then described the perioperative and oncologic outcomes of their series of RAN using a complete robotic approach. There is a lack of data comparing the perioperative and oncologic outcomes of the robotic-assisted and laparoscopic techniques, though. Herein, we evaluate the perioperative and oncologic outcomes of our initial cohort of RAN patients using matched comparisons to patients from our well-established practice of performing LN for UTUC.
Financial Disclosure: D.C.M. received grant funding from the Agency for Healthcare Research and Quality and the Urology Care Foundation/Astellas Rising Star in Urology Research Award. D.C.M. also received contract support from Blue Cross Blue Shield of Michigan for a position as Director of the Michigan Urological Surgery Improvement Collaborative. The remaining authors declare that they have no relevant financial interests. From the Department of Urology, University of Michigan Health System, Ann Arbor, MI Reprint requests: Sapan N. Ambani, M.D., Department of Urology, University of Michigan Health System, 500 East Medical Center Drive, TC 3875, Ann Arbor, MI 48109-5330. E-mail: [email protected] Submitted: January 31, 2013, accepted (with revisions): July 16, 2013
METHODS
ª 2014 Elsevier Inc. All Rights Reserved
We conducted a retrospective review of our prospectively maintained, institutional review board-approved, minimally invasive surgery database. All patients who underwent RAN from April 2009 to December 2011 were identified. RAN was performed using the da Vinci Robotic Surgical System (Intuitive Surgical, Inc., Sunnyvale, CA) by 1 of 3 attending surgeons (J.S.M., D.C.M., and A.Z.W.), each of whom completed 0090-4295/14/$36.00 http://dx.doi.org/10.1016/j.urology.2013.07.079
345
accredited urologic oncology fellowships. All patients treated with LN from January 2002 to December 2011 were identified, including those that were performed using hand-assisted and standard laparoscopic approaches. Previous literature has suggested that hand-assistance provides no difference in short-term outcomes with the only significant difference being shorter operative time.10 This similarity provided justification for inclusion of both approaches. All LNs were performed by 1 of 2 attending surgeons (J.S.W. and A.Z.W.), each of whom performed over 50 laparoscopic nephrectomies or nephroureterectomies before the beginning of this study period. Our techniques for hand-assisted and pure laparoscopic nephroureterectomy have been previously described.11,12 Our RAN technique is similar to that described by Eandi et al.9 In brief, the patient’s lower extremities are placed in the low lithotomy position, while their upper body is positioned in a partial flank position (60 angle to the table) and secured to the operating table using a beanbag and tape, with the ipsilateral arm supported on an arm board. The table is then rotated around the vertical axis 45 , and the robot is positioned over the patient’s ipsilateral shoulder to complete the renal and proximal ureteral dissection. For high-grade collecting systems and proximal ureteral tumors, the para-aortic or para-caval and hilar node dissections are performed in this position. Once the dissection is completed to the limit of this position, the robot is undocked, the table rotated back to the neutral position around the vertical axis, and the patient is rotated in the horizontal plane on the operating table to approximate a flat position. The robot is then repositioned between the patient’s legs and redocked to complete the distal dissection. Previously placed trocars are used with no additional ports required. The entire length of the ureter is dissected free and a self-closing clip is placed across the distal ureter to prevent upper tract urine spillage. The distal ureter is dissected through the intramural portion until the bladder wall is identified at the level of the ureteral orifice. We then enter the bladder sharply to circumferentially dissect out the entire orifice from the bladder. A previously placed urinary catheter as well as the pneumoperitoneum prevents urinary spillage from the bladder. The bladder is then closed in 2 layers in a running fashion using a 2-0 polyglactin suture cut to 8 inches. At the conclusion, the bladder is irrigated to ensure a water-tight closure. The distal para-aortic or para-caval, common iliac, external iliac, obturator, and internal iliac lymph nodes are then removed for highgrade distal ureteral tumors. Bipolar and monopolar cautery is used for hemostatic control for both LN and RAN. Hilar control is achieved using an Endo-GIA universal stapler (Autosuture, Covidien, Mansfield, MA) for both approaches e either en bloc or individually. Bipolar and monopolar cautery as well as selfclosing clips were used for hemostasis during the lymph node dissection. After nephroureterectomy, we routinely perform regular surveillance with history, physical examination, blood work, and imaging. We do not routinely use adjuvant intravesical therapy to prevent bladder recurrence, but rather initiate a protocol of bladder surveillance, including office cystoscopy at routine intervals. Patient demographics, pathologic findings, and perioperative data were obtained from our database and supplemented by our electronic medical records as needed. Comorbidities were characterized and stratified using the Charlson comorbidity index.13 Complications were classified according to severity as defined by Clavien.14 Matching was performed between LN and RAN in one-to-one fashion. Patients were matched by age 346
(10 years) and pathologic stage (TNM); those cases that could not be matched by these criteria were excluded. Statistical analysis was performed to compare operative and oncologic outcomes, using the paired t test for continuous variables and the McNemar test for categorical variables, with P
METHODS
RESULTS
CONCLUSION
To compare our initial robotic-assisted nephroureterectomy (RAN) experience with a wellestablished practice of performing laparoscopic nephroureterectomy (LN) to treat upper tract urothelial carcinoma (UTUC). We reviewed our prospectively maintained minimally invasive surgery database. Patients who underwent RAN from April 2009 to December 2011 were matched by pathologic tumor stage and age (10 years) to those who underwent LN. Twenty-two matched pairs were evaluated. Mean operative time (298 vs 251 minutes) and estimated blood loss (EBL, 380 vs 233 mL) were significantly higher for RAN, with a greater need for transfusion in this group. Complication rates were similar. The RAN group trended toward more frequent lymph node dissection and greater median node count when lymph node dissection was performed (59% vs 27% [P ¼ .07] and 5.5 vs 1.0 [P ¼ .13]). After a median followup of 10 months for RAN and 15 months for LN, no significant difference was seen in the rate of bladder (36% vs 37%) or distant (32% vs 23%) recurrence, with similar median time to any recurrence (9 months vs 4 months, P ¼ .32). RAN was associated with higher operative time and blood loss likely because of more frequent use of node dissection, robot repositioning, and technical inexperience. Lymph node dissection was more frequently performed with RAN, which reflects surgeon practice patterns. When a lymph node dissection was performed, the median node count was greater with RAN. Our initial experience with RAN suggests that this is an acceptable approach for the management of UTUC. UROLOGY 83: 345e349, 2014. 2014 Elsevier Inc.
pper tract urothelial carcinoma (UTUC) is rare, accounting for only 5% of all urothelial malignancies.1 As a result, the nuances of the appropriate management of this disease are not fully established. The historic standard management of UTUC is an open nephroureterectomy with excision of a bladder cuff to remove the entire urothelium of the affected renal unit. More recently, there has been an emphasis on using laparoscopic techniques to perform this procedure.2 Oncologic equivalence and improvement in postoperative convalescence has been shown when comparing laparoscopic nephroureterectomy (LN) with open nephroureterectomy,2-6 indicating that LN is an
U
effective treatment option. The advent of robotic technology has expanded the potential for the minimally invasive management of urologic conditions. Laparoscopic robotic-assisted nephroureterectomy (RAN) for UTUC was initially performed in combination with traditional laparoscopy.7,8 Eandi et al9 then described the perioperative and oncologic outcomes of their series of RAN using a complete robotic approach. There is a lack of data comparing the perioperative and oncologic outcomes of the robotic-assisted and laparoscopic techniques, though. Herein, we evaluate the perioperative and oncologic outcomes of our initial cohort of RAN patients using matched comparisons to patients from our well-established practice of performing LN for UTUC.
Financial Disclosure: D.C.M. received grant funding from the Agency for Healthcare Research and Quality and the Urology Care Foundation/Astellas Rising Star in Urology Research Award. D.C.M. also received contract support from Blue Cross Blue Shield of Michigan for a position as Director of the Michigan Urological Surgery Improvement Collaborative. The remaining authors declare that they have no relevant financial interests. From the Department of Urology, University of Michigan Health System, Ann Arbor, MI Reprint requests: Sapan N. Ambani, M.D., Department of Urology, University of Michigan Health System, 500 East Medical Center Drive, TC 3875, Ann Arbor, MI 48109-5330. E-mail: [email protected] Submitted: January 31, 2013, accepted (with revisions): July 16, 2013
METHODS
ª 2014 Elsevier Inc. All Rights Reserved
We conducted a retrospective review of our prospectively maintained, institutional review board-approved, minimally invasive surgery database. All patients who underwent RAN from April 2009 to December 2011 were identified. RAN was performed using the da Vinci Robotic Surgical System (Intuitive Surgical, Inc., Sunnyvale, CA) by 1 of 3 attending surgeons (J.S.M., D.C.M., and A.Z.W.), each of whom completed 0090-4295/14/$36.00 http://dx.doi.org/10.1016/j.urology.2013.07.079
345
accredited urologic oncology fellowships. All patients treated with LN from January 2002 to December 2011 were identified, including those that were performed using hand-assisted and standard laparoscopic approaches. Previous literature has suggested that hand-assistance provides no difference in short-term outcomes with the only significant difference being shorter operative time.10 This similarity provided justification for inclusion of both approaches. All LNs were performed by 1 of 2 attending surgeons (J.S.W. and A.Z.W.), each of whom performed over 50 laparoscopic nephrectomies or nephroureterectomies before the beginning of this study period. Our techniques for hand-assisted and pure laparoscopic nephroureterectomy have been previously described.11,12 Our RAN technique is similar to that described by Eandi et al.9 In brief, the patient’s lower extremities are placed in the low lithotomy position, while their upper body is positioned in a partial flank position (60 angle to the table) and secured to the operating table using a beanbag and tape, with the ipsilateral arm supported on an arm board. The table is then rotated around the vertical axis 45 , and the robot is positioned over the patient’s ipsilateral shoulder to complete the renal and proximal ureteral dissection. For high-grade collecting systems and proximal ureteral tumors, the para-aortic or para-caval and hilar node dissections are performed in this position. Once the dissection is completed to the limit of this position, the robot is undocked, the table rotated back to the neutral position around the vertical axis, and the patient is rotated in the horizontal plane on the operating table to approximate a flat position. The robot is then repositioned between the patient’s legs and redocked to complete the distal dissection. Previously placed trocars are used with no additional ports required. The entire length of the ureter is dissected free and a self-closing clip is placed across the distal ureter to prevent upper tract urine spillage. The distal ureter is dissected through the intramural portion until the bladder wall is identified at the level of the ureteral orifice. We then enter the bladder sharply to circumferentially dissect out the entire orifice from the bladder. A previously placed urinary catheter as well as the pneumoperitoneum prevents urinary spillage from the bladder. The bladder is then closed in 2 layers in a running fashion using a 2-0 polyglactin suture cut to 8 inches. At the conclusion, the bladder is irrigated to ensure a water-tight closure. The distal para-aortic or para-caval, common iliac, external iliac, obturator, and internal iliac lymph nodes are then removed for highgrade distal ureteral tumors. Bipolar and monopolar cautery is used for hemostatic control for both LN and RAN. Hilar control is achieved using an Endo-GIA universal stapler (Autosuture, Covidien, Mansfield, MA) for both approaches e either en bloc or individually. Bipolar and monopolar cautery as well as selfclosing clips were used for hemostasis during the lymph node dissection. After nephroureterectomy, we routinely perform regular surveillance with history, physical examination, blood work, and imaging. We do not routinely use adjuvant intravesical therapy to prevent bladder recurrence, but rather initiate a protocol of bladder surveillance, including office cystoscopy at routine intervals. Patient demographics, pathologic findings, and perioperative data were obtained from our database and supplemented by our electronic medical records as needed. Comorbidities were characterized and stratified using the Charlson comorbidity index.13 Complications were classified according to severity as defined by Clavien.14 Matching was performed between LN and RAN in one-to-one fashion. Patients were matched by age 346
(10 years) and pathologic stage (TNM); those cases that could not be matched by these criteria were excluded. Statistical analysis was performed to compare operative and oncologic outcomes, using the paired t test for continuous variables and the McNemar test for categorical variables, with P
