Robotic-assisted laparoscopic partial nephrectomy: surgical technique and clinical outcomes at 1 year


Henry Ho, Urology, Singapore General Hospital, Outram Road 169608, Singapore. e-mail:, or



To report our surgical technique of robotic-assisted laparoscopic partial nephrectomy (RLPN) for renal tumours of <7 cm and present their clinical outcomes, as minimally invasive PN is an increasingly viable option for small renal tumours.


From July 2005 to December 2006, 20 consecutive patients (mean age 58.2 years, sd 7.9) had RLPN and a follow-up of ≥1 year, all surgery being undertaken by one surgeon. All cases were elective except in one patient with a solitary kidney. We used the three-arm da Vinci robotic system (Intuitive Surgical, Sunnyvale, CA, USA) in a four-port, transperitoneal approach. Transient vascular occlusion was applied in all cases using a tourniquet technique. The tumour was excised with a 5-mm margin using cold-cut scissors, and the margins were assessed by frozen sections. The specimen was placed in an impervious bag for subsequent removal via the camera port. Under direct vision, we repaired all pelvicalyceal system entries with absorbable sutures. After the entire tumour bed surface was lined with FlosealTM (Baxter Healthcare, Deerfield, IL, ISA) the capsule/parenchyma was closed with running suture, reinforced by haemostatic clips.


The mean (sd) operative and warm ischaemia times were 82.7 (17.0) and 21.7 (2.4) min, respectively, and the mean estimated blood loss was 189 (32) mL. There were no intraoperative complications or conversion to open surgery. There was also no bleeding after surgery, perinephric haematoma or urinary leakage. The mean (sd) tumour size was 30.2 (2.4) mm, while margins were negative in all cases of malignancy. At the 1-year follow-up there was no local recurrence, renal functional deterioration or late surgical complications.


Our RLPN technique is a safe and feasible option for small renal tumours. Reproducible technique and good team co-ordination are pivotal for obtaining good oncological and surgical outcomes.


(open) (robotic) (laparoscopic) partial nephrectomy


(cold) (warm) ischaemic time


pelvicalyceal system




Asymptomatic small renal masses represent an increasingly common clinical scenario that every urologist inevitably encounters. In addition to preserving renal function, local tumour excision with negative surgical margins provides potentially curative treatment while obtaining histological confirmation of RCC [1]. Open partial nephrectomy (OPN) is the standard of care for renal tumours of <4 cm [2]. Minimally invasive approaches, e.g. pure laparoscopic PN (LPN) or robotic-assisted LPN (RLPN) are gaining ground as viable options, particularly in experienced high-volume centres [3,4].

While there are several reports on effective laparoscopic cooling techniques of the kidneys for prolonged cold ischaemic time (CIT) [5,6], its clinical acceptance remains limited. Thus, keeping the warm IT (WIT) brief is strictly observed for optimal renal function after surgery [7]. Current minimally invasive approaches (LPN/RLPN) focus on techniques to achieve this goal without compromising the oncological and surgical outcomes, particularly bleeding.

Since 2002, our centre has been using the three-arm Da Vinci robotic system (Intuitive Surgical, Sunnyvale, CA, USA) to assist in our LPN [8]. Its three-dimensional (3D) vision and wide-ranging Endo-wrist movements at the tip of the robotic arms permit easier intracorporeal suturing [9]. We have developed and simplified our RLPN technique; here we describe our approach for renal tumours of <7 cm and present the clinical outcomes for the most recent 20 patients with a 1-year follow-up.


From July 2005 to December 2006 we performed 20 consecutive RLPNs in our department of urology; the evaluation before RLPN included abdominal and chest contrast CT. All patients had contrast-enhancing tumours of <7 cm with an indeterminate radiological diagnosis or suspicion of malignancy. We included only the 20 most recent patients with a follow-up of ≥1 year. Patients with tumours of >7 cm or previous intra-abdominal surgery that precluded the transperitoneal approach were excluded. Tumours that required a WIT of >30 min are treated by OPN with cooling, and included endophytic tumours on the dorsal aspect in the upper pole of the kidney, which was difficult for our transperitoneal approach. All operations were performed by one surgeon (R.P.) experienced in laparoscopic and robotic urological surgery. The disease was staged by the uropathologist, according to the 1997 TNM classification system. Patients’ information was prospectively entered into a database. The operative duration was defined as being from the skin incision to its closure, which includes docking of the robot and console time. WIT began with placing an arterial tourniquet and ended with its release. Data were expressed as the mean (sd, range).

Under general anaesthesia, the patient is secured in the flank position with a 45° lateral tilt. The operating table is placed in hyperextension to improve the working space for the robotic ports (Fig. 1); we do not use retrograde access or a ureteric catheter. We use the transperitoneal approach in all our RLPNs. A peri-umbilical incision is made and pneumoperitoneum is created with the Veress needle technique. The first port is a cutting 12-mm port (Ethicon Endo-surgery Endo-Surgery Inc, Cincinnati, OH, USA) for the camera system. The two robotic ports (8 mm) and one assistant port (12-mm Ethicon) are inserted under direct vision, as shown in Fig. 2. We maintain the pneumoperitoneum at 12–15 cmH2O throughout the procedure.

Figure 1.

The patient’s position for RLPN; supported flank position with a 45° lateral tilt with the operating table placed in hyperextension.

Figure 2.

Port positions: peri-umbilical camera port. The landmark for the upper robotic port is 10 cm from the subcostal margin and the lower robotic port is placed at the midpoint on an imaginary line between the antero-superior iliac spine and the umbilical port. The assistant port is inserted midway between the camera and the upper robotic port.

After creating the pneumoperitoneum the da Vinci robot is docked at the patient’s back. The robotic arms go round the patient’s side and instruments are inserted from the front. The docking time, which includes linking of the robot to its ports, and insertion of the bipolar scissors and grasper for the robotic arms, is <10 min. We use the 30°-down lens for the entire procedure. The assistant surgeon is positioned at the patient’s side, and is responsible for changing the robotic arm instruments and assists via the assistant-port with liver retraction (right-sided tumour), suction, insertion of sutures, placing Hem-o-Lok® clips (Weck Closure Systems, Research Triangle Park, NC, USA) and applying FloSealTM (Baxter Healthcare, Deerfield, IN, USA).

After the overlying colon has been mobilized, the next step is hilar control, for which we use the tourniquet technique [10–12]. The renal artery and vein are dissected sufficiently for a tourniquet device to be looped around each of them. When transient vascular occlusion is required, the plastic tube slides towards the renal vessel and a Hem-o-Lok clip is placed on the vascular loop next to the tube. The clip serves as a stopper to maintain the tourniquet around the renal vessels. To facilitate identification, red and blue loops are used for artery and vein, respectively. Intravenous mannitol (20%, 200 mL) infusion is completed before the arterial tourniquet is applied. First, we usually occlude the arterial loop before the venous loop and the WIT is recorded.

Next, the perinephric fat is freed from the kidney capsule except for that overlying the tumour. It is important that the subcapsular plane of the kidney is not entered. The incisional boundary (5 mm from the tumour margin) is marked on the capsule with monopolar cautery. For endophytic tumours with concealed boundaries, we used a laparoscopic ultrasound probe to clearly delineate its boundary. After we have changed one of the robotic arms to a grasper and the other to cold-cut scissors, the tumour is excised, respecting the principles of open surgery (Fig. 3). In an impervious specimen bag (EndoCatchTM bag device, Autosuture, Norwalk, CO, USA) the specimen is placed away from the operative site. Individually labelled margins, taken from the four quadrants of the tumour bed, are examined for tumour involvement by frozen-section analysis.

Figure 3.

Tumour excision with cold-cut scissors and grasper.

Once the tumour has been excised, both robotic arms are changed to needle drivers. Next, the tumour bed is inspected for any major bleeding or truncated ends of vascular branches. Severe bleeding can temporarily be controlled by further tensioning on the vascular tourniquet or direct suture ligation of bleeding vessels. The latter requires the assistant to provide appropriate exposure of the tumour bed. Generally, suturing is done with absorbable sutures (polyglactin 3/0). Any non-bleeding truncated ends of blood vessels are similarly closed.

Under direct 3D vision we identify all entries into the pelvicalyceal system (PCS) and repair them with absorbable sutures (polyglactin 4/0). We do not use laparoscopic clips to assist in its closure, or a retrograde ureteric catheter to inject methylene blue to guide the repair.

Next, the entire surface of the tumour bed is covered with FloSeal, applied by the assistant surgeon using a long-tipped applicator. Our technique for parenchymal closure is based on the application of Hem-o-Lok clips requiring minimal knot tying [12–14]. The suture (polyglactin 1/0) is 20-cm long with a Hem-o-Lok clip at its terminal end. A stabilizing knot is prepared beyond the clip, serving as a stopper to prevent its slippage. Each parenchymal bite includes the capsule ≈1 cm from each edge of the tumour bed. After the suture has reached the corresponding opposite side of the defect, another Hem-o-Lok is placed on the suture, next to the surface of the kidney (Fig. 4). Tension can be gradually increased by sliding the Hem-o-Lok along the suture. This is repeated until the defect is closed completely. Next, the vascular tourniquets are released, with vein followed by the artery. This is achieved by cutting the vascular loop or unlocking the Hem-o-Lok clip by twisting both arms in the opposite direction.

Figure 4.

Capsular/parenchymal closure with suture and Hem-o-Lok clips; no surgical bolster was used.

Gerota’s fascia is closed and the colon is re-approximated to its original position, overlying the kidney. A non-suction drain is placed near the operative site to monitor extravasation or bleeding. The specimen is removed via the camera port by enlarging the port to 2–3 cm.

The patients are closely monitored overnight, including a full blood count and renal function tests. They resume oral intake on the first day after surgery. The drain is removed when its drainage is minimal. We assess the operated kidney using ultrasonography on the fourth day after surgery to detect any perinephric collection (blood, fluid or urine). Before discharge from the hospital, the renal function test is repeated.

During the follow-up we use kidney ultrasonography at 3 months and contrast CT of the kidneys biannually for 2 years. The serum creatinine level is assessed during every outpatient visit.


The patients’ demographics and preoperative tumour characteristics are presented in Table 1. There was no case of multiple renal tumours in the same patient; no patient had any previous renal intervention. The data during and after surgery are also shown in Table 1. Transient vascular occlusion was by the tourniquet technique in all cases. Five kidneys had more than one artery or vein, while we had no additional problems using this technique. There were eight cases that needed closure of the PCS, and their mean WIT was 24.3 min There was no case where we were unable to close the defect, including the 5.5 cm tumour. There were no complications after surgery, e.g. bleeding, haematoma formation, urinary leak or ileus.

Table 1. 
The demographic and operative data of the 20 patients
VariableMean (sd, range) or number
Age, years58.2 (7.9, 36–78)
Gender (male/female)13/7
Preop. serum creatinine, mg/dL0.89 (1.81, 0.68–1.05)
Indications for RLPN 
 Imperative1 (solitary kidney)
Normal contralateral kidneys19
Tumour size on CT, mm35.3 (4.8, 20–55)
Tumour side (left/right)11/9
Tumour classification 
Tumour location 
 upper pole6
 lower pole6
Operative duration, min82.8 (17.0, 75–95)
WIT, min 
 overall21.7 (2.4, 15–27)
 no PCS closure19.8 (1.3, 15–21)
 PCS closure (8)24.3 (3.0, 22–27)
Estimated blood loss, mL189 (32, 50–260)
Transfusion rate0
Urinary leak0
Perinephric haematoma0
Hospital stay, days4.8 (1.3, 4–7)
Tumour size, mm30.2 (2.4, 20–55)
Positive surgical margins0
 Clear cell11
RCC grade 
RCC stage 
Benign tumours7

Table 1 also lists the pathological results; there was primary RCC in 65% of patients, with negative surgical margins in all patients. The single case of grade 3 RCC was also the largest tumour in the series. Given the margin status, there was no local recurrence. All the patients had a follow-up of ≥1 year; there was no elevated serum creatinine level, local tumour recurrence or distant relapse, based on the 1 year CT, and no delayed surgical complications such as incisional hernia.


PN is the standard of care for renal tumours of <4 cm (stage T1a) [1,4], and LPN has progressively been shown to be a safe alternative with comparable oncological and surgical outcomes to OPN [3]. However, being a technically demanding procedure that requires advanced laparoscopic skills, it is limited to some high-volume centres. Moreover, the risk of postoperative renal impairment is directly related to the duration of WIT. A multi-institutional study of OPN in solitary kidneys showed that the WIT and CIT should be limited to 20 and 35 min, respectively, to avoid irreversible renal complications [7]. Until laparoscopic renal cooling is widely used, minimally invasive PN will continue to focus on those techniques that reduce WIT, i.e. simplified suturing techniques or robotic assistance.

Simplified techniques without compromising the clinical outcome are of paramount importance for both LPN and RLPN. In the most ideal setting, every available haemostatic technique should be used to prevent bleeding. Unfortunately, time management is essential in PN. As such, we need to streamline the most difficult steps to achieve this goal. With our current technique for a mean tumour size of 30.2 mm, our mean operative duration and WIT were 82.7 and 21.7 min, respectively. Our early and regular radiological surveillance detected no bleeding complications or urine leakage.

To our knowledge, we were the first centre to report the use of the da Vinci robotic system to assist in LPN [8]. Since then, we have done more than 50 cases and simplified our technique. The surgical team consisted of a primary surgeon (R.P.) and assistant surgeons, who have laparoscopic knowledge and skills. This ensured understanding and co-ordination between the console and assistant surgeons. In the present report we included only the most recent 20 patients (after we had gained full training and experience) with a 1-year follow-up.

We only use the transperitoneal approach, with its larger working space for the robotic arms and familiar landmarks to facilitate dissection and preparation. In our first report, we adopted the transperitoneal and retroperitoneal approaches for anterior and posterior/lateral tumours, respectively. The space limitation in the latter approach frequently impeded the use of robotic assistance [8]. Unfortunately, the transperitoneal approach also makes it difficult to use RLPN for tumours in the dorsal part of the upper pole. As such, we choose OPN for these tumours, particularly if they are endophytic. Conventional gastrointestinal concerns, such as bowel injury and postoperative ileus, remain issues for the transperitoneal approach, although we have not encountered them.

One important technical point is that the kidney capsule must be maintained, particularly during denuding of the perinephric fat. If the subcapsular plane is entered, bleeding can be substantial. This will impair visibility, leading to inaccurate dissection and positive margins. Moreover, the fibrous renal capsule is crucial for stabilizing the suture with its Hem-o-Lok clips during subsequent parenchymal reconstruction.

We are the only centre using the tourniquet technique [10] for transient vascular control amongst those using RLPN [15–17]. This approach ensures that vascular control remains with the console surgeon. The elastic vascular loops allow selective control of the renal vessels and avoid intimal trauma. In the other techniques, e.g. laparoscopic bulldog and Satinsky clamps, vascular control is not with the console surgeon, as they are applied by the assistant surgeon. In the former, the need for perfect alignment with the laparoscopic applicator before its removal or adjustment makes it less user-friendly in a time-conscious environment. The laparoscopic Satinsky clamp is another option for en-bloc renal artery and vein control. When in use, it occupies a port and can be traumatic to the vascular endothelium [18]. It also lacks the capability of selective renal vessel control, which was suggested to improve the preservation of renal function [19].

The 3D magnified vision for the console surgeon enhanced the ability to detect any PCS entry. In a retrospective comparison of 59 patients, there was no difference in the incidence of urinary fistula between the groups with and without retrograde injection of methylene blue via a ureteric catheter. They concluded that visual cues can be used to identify most PCS entries [20]. We concurred with their conclusion and observed similar capabilities with the excellent 3D vision offered by the da Vinci robotic system. Although we would not know the true incidence of PCS entries, from our early and regular radiological investigations, there was no case of urinoma.

The eight patients that needed closure of the PCS had a slightly longer mean WIT than those without (19.8 vs 24.3 min). Desai et al.[21] reported similar findings in their LPN series, but our mean WIT of those who needed PCS closure was shorter (24.3 vs 30.2 min). We also used absorbable sutures instead of laparoscopic clips for PCS closure, which eliminated the risk of erosion of the collecting system [12]. These are the likely benefits of facilitated intracorporeal suturing with the da Vinci robotic system.

FloSeal is an effective haemostatic adjunct for minimally invasive PN. It is a biosynthetic tissue sealant, which consists of thrombin/gelatine granules and aids in haemostasis. It is easily applied laparoscopically to the tumour bed and is activated by blood [22]. Desai et al.[23] reported its use in a porcine model of hand-assisted LPN and there was no haematoma after surgery. In a clinical study of 68 LPN that compared the use of FloSeal or no sealant, the former group had lower overall complication rates (15.9% vs 36.6%; P < 0.001) and haemorrhagic complications (3.2% vs 11.8%; P = 0.08) [24]. Unfortunately, FloSeal activation is hindered by contact with urine [25] and therefore it is imperative that any breach in the PCS is closed before it is applied.

Our renorraphy technique did not include a classical surgical bolster, a popular haemostatic technique [12,26]. Weight et al.[27] reported LPN without bolstering in 23 patients, and their selection criteria were based on an intraoperative surgical assessment of the defect. They reported a shorter WIT with comparable complication rates. We have concerns about feasibility if the tumour bed defect is too wide and shallow for direct closure. However, we have not yet encountered such a case. As commented by Agarwal et al.[14] the surgical bolster might be for added security, but was not fundamental for haemostatic closure.

Our mean operative duration and WIT were 82.7 and 21.5 min, respectively, which is comparable with other laparoscopic or robotic series [15,16]. In the largest series of LPN, the mean WIT was 30.7 min [3]. Caruso et al.[15] reported their single centre non-randomized comparison between RLPN and LPN; with 10 cases in each arm they concluded that there was no statistically significant difference in the patients’ outcomes during and after surgery. The average operative duration for RLPN was 279 min. Their haemostatic and reconstructive technique included defect-base cauterization with TissueLink, Gelform packing, fibrin glue injection and mattress suture closure over Surgicel bolsters. Our mean operative duration was shorter than other RLPN series, which exceed 120 min. The other reports represent initial surgical experience, and the use of many haemostatic techniques might also contribute to the longer operative and WIT [17]. We think that our simplified technique and consistent team co-ordination are important to achieve our current outcomes.

Being a retrospective study, there is no control arm with a different technique for comparison of ischaemic times or surgical outcomes. We also acknowledge that the serum creatinine level is an insensitive test for renal function, particularly in setting of a normal contralateral kidney. Radionuclide studies to determine the residual function in the operated kidney will be useful in a research setting [28]. Unfortunately, in our clinical setting, its use is limited.

In addition to the high cost associated with the da Vinci robotic system, RLPN still requires advanced urological laparoscopy with considerable training and experience. Our technique shows that it is possible to have a short operative duration and WIT with RLPN. Further comparative studies between RLPN and LPN will determine if robotic assistance reduces the required training and experience. For all the technological advantages of RLPN, the lack of tactile feedback remains a disadvantage. However, intraoperative visual cues are usually sufficient to compensate.

In conclusion, RLPN is a safe and feasible option in the management of selected small renal tumours. Our technique has a shorter operative duration and WIT with no early complications, compared with other series of RLPN. At the 1-year follow-up, patients had no local tumour recurrence. Reproducible technique and good team co-ordination are pivotal in achieving reliable results. In centres already using robotic-assisted radical prostatectomy, RLPN will simply be a natural extension; in the others, it might be an attractive option.


None declared.