Perioperative outcomes of robot-assisted nephroureterectomy for upper urinary tract urothelial carcinoma: a multi-institutional series


Correspondence: Joseph Pugh, Department of Urology, 1600 SW Archer Rd, PO Box 100247, Room N-203, Gainesville, FL 32610, USA.




  • To review a multi-institutional series of robot-assisted nephroureterectomy (RANU) for management of upper urinary tract urothelial carcinoma (UUTUC) with respect to technique and perioperative outcomes.

Patients and Methods

  • Between May 2007 and July 2011, 43 RANU were performed at three institutions for UUTUC with review of perioperative outcomes.
  • A three- or four-armed robotic technique was used in all cases based on surgeon preference and the entirety of all procedures was performed using the robot-assisted technique.
  • Single and two robot-docking techniques are described.


  • The mean (range) operating time was 247 (128–390) min, blood loss was 131 (10–500) mL and the median (range) length of stay was 3 (2–87) days.
  • Pathology was pTa in nine patients, pT1 in 14 patients, pT2 in three patients, pT3 in 15 patients and pT4 in two patients.
  • Lymph node dissection was performed in 22 patients (51%) with a mean (range) lymph node count of 11 (4–23).
  • There were six postoperative complications: bleeding requiring a blood transfusion (grade II), splenic bleeding (grade IV), two cases of pneumonia (grade II) and two cases of rhabdomyolysis (grades II and IV).
  • Nine recurrences (six bladder, two within the retroperitoneum and one in the contralateral collecting system) have been found to date on routine surveillance with a mean follow-up of 9 months.


  • RANU is a feasible alternative to laparoscopic and open techniques.
  • Particular steps of the operation including sutured closure of the cystotomy and regional lymphadenectomy are facilitated with the use of robot-assisted surgery.
  • Long-term outcomes are necessary to assess the relative efficacy of these approaches to more established techniques; however, early perioperative outcomes appear promising.


Upper urinary tract urothelial carcinoma (UUTUC) accounts for 5% of all urothelial tumours [1]. The ‘gold standard’ treatment remains radical nephroureterectomy (NU) with resection of the bladder cuff [2]. Various surgical approaches to NU have been described, including open, laparoscopic, hand-assisted laparoscopic and robot-assisted techniques. Laparoscopic NU (LNU) was first described by Clayman et al. [3] in 1991. Since then, several series have noted comparable oncological outcomes as compared with the open procedure with the added benefit of decrease length of stay, analgesic requirement, and blood loss [4-8]. Many techniques have been described for management of the intravesical ureter and bladder cuff during LNU. Options include transvesical, intravesical and endoscopic approaches; however, there is no consensus on the best approach [9].

With the recent expansion of robot-assisted surgery in the field of urology and following the success of robot-assisted prostatectomy and partial nephrectomy, robot-assisted surgery has being applied to many other genitourinary diseases [10-13]. Herein, we present a detailed description of one vs two robot docking for robot-assisted NU (RANU) and perioperative outcomes of our multi-institutional experience.

Patients and Methods

After Institutional Review Board approval, a retrospective chart review was performed at three large tertiary care academic institutions, University of Florida, New York University and Wake Forest University, to identify patients who underwent RANU for the treatment of UUTUC from May 2007 to July 2011 by four surgeons. All patients underwent appropriate diagnostic imaging and evaluation including history and physical examination, cystoscopy, routine serum chemistries and cross-sectional imaging with CT or MRI. Patients with clinical ≤T2 disease were considered for RANU. The use of a single vs two robot docking technique was based upon surgeon preference and will be described herein.

Single Robot-docking Technique

The technique of single docking RANU has been previously described by Hemal et al. [14]. In brief, the patient is positioned in a modified 45 ° flank position with the diseased side up. The table is flexed and the patient is put in slight Trendelenburg of ≈15 °. A 12-mm camera port is placed at the level of the umbilicus, lateral to the rectus sheath. Three 8-mm robotic ports are then placed and used interchangeably during different parts of the operation. The first robotic port is placed lateral to the rectus sheath and 7–8 cm cranial to the camera port. A second robotic port is placed lateral to the rectus sheath and 7–8 cm caudal to the camera port. A third robotic port is placed ≈5 cm cranial to the iliac crest, in the anterior axillary line or slightly medial to it. A 12-mm assistant port is placed within 2–3 cm cranial or caudal of the umbilicus. The robot is then docked perpendicular to the patient over the back. Either a 0 ° or 30 ° down lens is used for the nephrectomy and lymphadenectomy portion of the procedure depending on patient anatomy. The first robotic port is the left arm (monopolar curved scissors), the second robotic port is the right arm (Maryland bipolar forceps) and the third robotic port is the accessory arm to assist with retraction (ProGrasp forceps). After completion of the nephrectomy and lymphadenectomy, the camera with a 30 ° down lens is directed towards the bladder. The Maryland bipolar forceps are moved to the third robotic port, the monopolar scissors are moved to the second robotic port and the ProGrasp forceps are moved to the first robotic port. The distal ureteric dissection, bladder cuff excision and cystotomy closure are then performed.

Two Robot-docking Technique

The two robot docking technique has been previously described by Willis et al. [15]. Patient positioning is similar to the one docking description above, using a modified flank position. A total of four ports are used (Figs 1, 2). Port placement begins with a 12-mm paraumbilical camera port. One 8-mm robotic port is then placed lateral to the rectus muscle near the anterior axillary line just below the level of the umbilicus. A second 8-mm robotic port is placed 3–5 cm below the costal margin lateral to the rectus muscle. These ports accommodate the left and right robotic arms (Maryland bipolar forceps and curved monopolar) for the nephrectomy and lymphadenectomy portion of the operation, respectively. For the surgical assistant, a 12-mm port is placed in the midline midway between the umbilicus and pubis. In obese patients with a large abdominal pannus, this port configuration may require a slight lateral shift toward the kidney to allow for optimal visualisation and to reach to the target organ. A 30 ° down lens is used for the entire procedure. The robot is then docked at 45 ° above the patient's shoulder using the three upper ports for the nephrectomy and lymphadenectomy portion of the procedure. Once complete, the robot is re-docked at 45 ° above the patient's hip to accomplish the distal ureterectomy and bladder cuff dissection using the lower three set of ports. The 12-mm assistant port is exchanged for an 8–13 mm convertible robotic port (Intuitive Surgical, Sunnyville, CA, USA) converted to an 8-mm robotic port. This unique robotic port allows for an 8-mm cannula to intussucept within a 13-mm outer sheath through a connector. The convertible robotic port and the lower 8-mm robotic port accommodate the left and right robotic arms (Maryland bipolar forceps and curved monopolar) for the distal ureteric dissection and bladder cuff excision. Consequently, the assistant uses 5-mm instrumentation through the subcostal 8-mm robotic port by capping the 5-mm adapter.

Figure 1.

Trocar configuration for nephrectomy portion.

Figure 2.

Trocar configuration for distal portion.

In all cases, the bladder cuff and intravesical ureter was managed via direct excision of the ureterovesical junction and closed in a two-layered sutured repair adhering to open principles. As such a midline sagittal cystotomy was avoided. Regional perihilar and pelvic lymphadenectomy was performed based on clinical stage, grade, preoperative imaging characteristics and intraoperative findings.


Table 1 lists patient demographics and perioperative data for our cohort. In all, 43 patients from the three institutions were analysed. The mean (range) age was 68.3 (51–92) years and operating time was 249 (128–390) min, the median (range) length of stay was 3 (2–87) days. There were six complications in this group including one case of postoperative bleeding requiring blood transfusion (Grade II), an emergent take back for splenectomy secondary to postoperative splenic bleeding (Grade IV), two case of pneumonia (Grade II) and two cases of transient rhabdomyolysis (Grade II and Grade IV). Table 2 lists the pathological finding and short-term follow up. The two patients with pT4 disease were upstaged after surgery and one had a positive margin. Lymph node dissections were performed in 22 of the 43 patients (51%). For patients who underwent regional lymph node dissection the mean (range) lymph node count was 11 (4–23). Adjuvant chemotherapy was offered to all patients with advanced stage disease and given to six of the 15 patient with T3 disease and both patients with T4 disease. The mean (range) follow-up for RANU was 9.6 (2–36) months. Nine patients had recurrences on routine surveillance of which six recurred within the bladder, two within the retroperitoneum both in patients with high grade pT3 disease, and one in the contralateral collecting system.

Table 1. Baseline demographics and perioperative RANU data
Number of patients43
Male: female29:14
Mean (range) age, years68.3 (51–92)
Side, n18 right, 25 left
Mean (range) operating time, min247 (128–390)
Mean (range) estimated blood loss, mL131 (10–500)
Median (range) hospital stay, days3 (2–87)
Complications, n (%)6 (14)
Table 2. Pathological findings of RANU
VariableN (%) or mean (range)
Tumour stage: 
pTa9 (20)
pT114 (33)
pT23 (7)
PT315 (35)
pT42 (5)
High33 (77)
Low10 (23)
Surgical margins: 
Positive1 (2)
Negative42 (98)
Lymph nodes: 
Positive4 (9)
Negative18 (42)
Not performed21 (49)
Mean (range) follow-up, months9.6 (2–36)
Total recurrences9 (21)
Bladder6 (14)
Local/retroperitoneal3 (7)


With the recent expansion of robot-assisted surgery in the field of urology, the robotic platform has being applied to treat many genitourinary diseases. Robot-assisted prostatectomy is a well-accepted treatment method for localised prostate cancer with excellent oncological outcomes [16]. Robot-assisted surgical techniques are currently been used to treat various bladder and kidney cancers [17-19]. For UUTUC, the ‘gold standard’ remains NU. RANU mimics the open technique by formal and direct excision the bladder cuff.

For the robot-assisted technique in general, it is the authors' experience that the robot was most useful duringthe regional lymphadenectomy and for suturing of the cystotomy after excision of the distal ureter. Single or two docking techniques have been described for RANU [14, 15]. Each technique has advantages and disadvantages that may affect surgeon preference. The single docking technique may save time by eliminating the time to re-dock. Cases done with a single docking technique were on average 49 min shorter than the two docking technique. However, with a skilled robotics team, redocking the robot can be done in <10 min. One disadvantage of using the single docking technique is the need for a total of five (three robotic, one camera and one assistant) ports vs four for the two docking technique (two robotic, one camera and one assistant/convertible). The two docking technique may ease in the dissection of the distal ureter and bladder cuff excision due to a more favourable trocar position for dissection within the pelvis. A disadvantage of this technique is the assistant port is converted to a robotic 8-mm port with the second docking, resulting in having the assistant use only 5-mm ancillary instruments passed through the 5-mm adapter. Although this did not pose a problem when using the suction-irrigator or small hemoclips, sutures had to be passed into the peritoneal cavity through the 8–13 mm convertible robotic port. Taken together, the choice of either operative setup is based on surgeon discretion and experience.

Several randomised control trials have been performed comparing the efficacy of LNU to the open technique [20, 21]. These studies reported significant advantages with the laparoscopic approach and confirmed comparable oncological outcomes with open surgery. Hand-assisted laparoscopic techniques have also been described as well as a robot-assisted approach introduced by Park et al. [22]. Table 3 [14, 22-25] lists published series of minimally invasive NU; however, to date, there are no randomised controlled trials comparing these minimally invasive approaches.

Table 3. Comparison of NU techniques
ReferenceTechniqueNOperating time, minEstimated blood loss, mLLength of stay, days
Hemal et al. 2011 [14]RANU151841033
Eandi et al. 2010 [23]RANU113262005
Park et al. 2009 [22]RANU112471067
Berger et al. 2008 [25]LNU1001822484
Wolf et al. 2005 [24]Hand-assisted NU532793304
Present seriesRANU432491333

When comparing the present series with Park et al. [22], which used a similar two robot-docking technique, our operating time and estimated blood loss were similar. Hospital stay was shorter in the present series (i.e. 5 vs 7 days), which most probably reflects differences in practice patterns between the two institutions. We also had comparable outcomes to other published RANU series by Hemal et al. [14] and Eandi et al. [23] for operating time, blood loss and length of stay. Our operating time and blood loss was also slightly lower than that of the published series of hand-assisted NU by Wolfe et al. [24]. When comparing the present RANU with the LNU of Berger et al. [25], operating time was longer and blood loss was less with the robot-assisted approach. However, comparing these studies is confounded by the number of patients who underwent lymph node dissection (18 of 100, 18%) as compared with 51% (22 of 43) of patients in the present study. Regional lymph node dissections were done in high-risk patients in the present cohort at the discretion of the surgeon. This was based on clinical stage, high-grade disease, preoperative imaging characteristics and/or intraoperative findings. A regional lymph node dissection can provide better disease staging and may be curative in patients with limited nodal disease [26].

Although there were few complications in the present series, there were two patients in the present cohort that developed transient rhabdomyolysis after the procedure that warrant further discussion. Both patients were morbidly obese with a body mass index (BMI) of >40 kg/m2. Patients with morbid obesity pose unique challenges irrespective of surgical approach whether open or by minimally invasive including: neuromuscular compromise, increased operating time, wound infection and pulmonary complications. Despite extensive and appropriate preoperative planning with adequate padding of all pressure points, their extreme BMI in addition to the duration of surgery (one case lasting 330 min) probably predisposed them to postoperative rhabdomyolysis. One patient required temporary haemodialysis but eventually both recovered fully with return of normal renal function. No other sequalea from this complication was noted during their hospital course or follow-up. Of note, rhabdomyolysis after laparoscopic nephrectomy is a rare complication that has been previously reported by Glassman et al. [27]. Several risk factors have been shown to increase the risk of developing rhabdomyolysis including high BMI, prolonged operating time, transient intraoperative hypotension and the lateral decubitus position. It is important to have a high clinical suspicious before and after surgery, as early recognition is paramount to full recovery [28].

There are several limitations of the present study that should be addressed. The present study represents a multi-institutional retrospective case series. As such only limited conclusions can be drawn for comparative effectiveness with other treatments. Certainly a prospective randomised control study comparing robot-assisted techniques with open, laparoscopic or hand-assisted laparoscopic techniques would be the optimal study design to assess the clinical efficacy and cost comparison between these methods. Another limitation of the present study is the relatively short follow-up. The mean follow-up in the present cohort was ≈9 months. Long-term oncological outcomes will be a subject of a future analysis of our series. However, despite these limitations, robot-assisted management of UUTUC in this early analysis appears comparable with currently available published outcomes of minimally invasive NU.

In conclusion, RANU is a feasible alternative to other established minimally invasive surgical techniques for treating UUTUC. Certain advantages may include ease of excision of intravesical ureter, suturing for cystotomy closure and lymph node dissection. These techniques may be especially useful for surgeons with limited laparoscopic training or for more experienced teams who endeavour to expand their robotics offerings to surgery of the UUT. Nevertheless, additional studies are needed to evaluate the long-term efficacy of these procedures compared to the laparoscopic and open techniques including cost considerations.

Conflict of Interest

None declared.


body mass index


(laparoscopic) (robot-assisted) nephroureterectomy


(upper urinary tract) urothelial carcinoma