Robotic and laparoscopic partial nephrectomy: a matched-pair comparison from a high-volume centre

Authors

  • Monish Aron,

    1. Center for Laparoscopic and Robotic Urology, Department of Urology, Glickman Urological and Kidney Institute, Cleveland Clinic, Cleveland, OH, USA
    Search for more papers by this author
  • Phillipe Koenig,

    1. Center for Laparoscopic and Robotic Urology, Department of Urology, Glickman Urological and Kidney Institute, Cleveland Clinic, Cleveland, OH, USA
    Search for more papers by this author
  • Jihad H. Kaouk,

    1. Center for Laparoscopic and Robotic Urology, Department of Urology, Glickman Urological and Kidney Institute, Cleveland Clinic, Cleveland, OH, USA
    Search for more papers by this author
  • Mike M. Nguyen,

    1. Center for Laparoscopic and Robotic Urology, Department of Urology, Glickman Urological and Kidney Institute, Cleveland Clinic, Cleveland, OH, USA
    Search for more papers by this author
  • Mihir M. Desai,

    1. Center for Laparoscopic and Robotic Urology, Department of Urology, Glickman Urological and Kidney Institute, Cleveland Clinic, Cleveland, OH, USA
    Search for more papers by this author
  • Inderbir S. Gill

    1. Center for Laparoscopic and Robotic Urology, Department of Urology, Glickman Urological and Kidney Institute, Cleveland Clinic, Cleveland, OH, USA
    Search for more papers by this author

Monish Aron, Glickman Urological and Kidney Institute, Cleveland Clinic, 9500 Euclid Ave/A100, Cleveland, OH 44195, USA.
e-mail: aronm@ccf.org

Abstract

OBJECTIVES

To evaluate the relative merits of robotically assisted partial nephrectomy (RPN), using a matched-pair analysis, with laparoscopic PN (LPN).

PATIENTS AND METHODS

Between July 2006 and August 2007, 12 patients had RPN for tumour; the outcomes were compared retrospectively with 12 matched patients who had LPN. Patients were matched for age, gender, body mass index, American Society of Anesthesiologists score, tumour side, size and location, and the specific technique used (early vs conventional unclamping). Operative measures evaluated included operative time, estimated blood loss, warm ischaemia time (WIT), and number of ports used. Outcomes measured included serum creatinine and estimated glomerular filtration rate before and after surgery, length of hospital stay, transfusion rate, operative and 30-day complication rate, and surgical margin status.

RESULTS

Overall there were no differences in perioperative variables (WIT, estimated blood loss, surgery time, length of stay) between the groups. Fewer ports were used during LPN. Renal functional outcomes, transfusion rate and complication rates were comparable. Two RPN cases required conversion to standard LPN. A subset analysis of six patients in each group who had early unclamping showed a 7-min shorter WIT with LPN (14 vs 21 min, P = 0.05), despite larger tumours being treated with LPN (3 vs 2.4 cm, P < 0.01) in this subset.

CONCLUSIONS

RPN is a developing procedure, and is technically feasible and safe, albeit with a longer warm WIT than LPN. Further experience is necessary to determine the relative merits of RPN.

Abbreviations
(R)(L)(O)PN

(robotically assisted) (laparoscopic) (open) partial nephrectomy

BMI

body mass index

ASA

American Society of Anesthesiologists

MINSS

minimally invasive nephron-sparing surgery

FSA

frozen-section analysis

WIT

warm ischaemia time.

INTRODUCTION

The incidence of renal cancer increased annually by nearly 2–4% between 1975 and 1995 [1,2]. The greatest increase in incidence was for localized tumours, attributed in part to a significant increase in the use of cross-sectional abdominal imaging [1]. At present, of all renal tumours, 48–66% are small, localized and incidentally detected [3]. Up to 38% of renal tumours surgically excised in contemporary practice are ≤4 cm [4]. Despite recent developments in probe-ablative therapies, surgical excision remains the cornerstone of treatment for RCC.

Open partial nephrectomy (OPN) has been the reference standard for managing the small renal mass, with the potential benefit of better renal functional outcomes [5,6] and equivalent cancer control to radical nephrectomy [7–10]. However, OPN entails the morbidity of the muscle-cutting, open-flank incision, which can be durable in up to half of patients [11]. In addition, substantially many (≈30%) of small solid renal masses are benign [12]. The clear and present need to decrease procedure-related morbidity has fuelled significant advances in minimally invasive nephron-sparing surgery (MINSS) in the last 5 years.

For extirpative MINSS, the reference standard is laparoscopic PN (LPN); this involves intricate tumour excision and precise suturing in a time-sensitive manner. With the stated intent of facilitating intracorporeal suturing robotically assisted PN (RPN) has been explored by a few authors, and remains under development [13–17]. At our institution, our experience with LPN is robust, with >800 contemporary patients; as such, we sought to evaluate the relative merits of RPN from a team with wide experience of LPN.

PATIENTS AND METHODS

Between July 2006 and August 2007, 12 selected patients had RPN for a single small unilateral renal mass. The perioperative and 3-month functional outcomes were compared with those of 12 matched patients treated with LPN in the same period. Patients were matched retrospectively on the basis of age (within 10 years), gender, body mass index (BMI; within 5 points), American Society of Anesthesiologists (ASA) score, tumour side, size (within 10 mm), and location (upper, middle or lower pole), and the specific technique used (early hilar unclamping vs conventional unclamping). Patients were matched for the clamping technique to ensure comparability between the groups, as the warm ischaemia time (WIT) differs substantially between the early and conventional unclamping techniques.

It was not possible to match these two groups on the basis of the degree of technical difficulty in each patient, as this is a subjective intraoperative assessment by the surgeon. However, as this was our initial experience with RPN, hilar tumours and completely intraparenchymal tumours were excluded from robotic surgery.

Summary statistics were calculated and matched-pairs analysed using appropriate software, with P < 0.05 taken to indicate significant differences.

The LPN technique was detailed previously [18]; essential steps include strategic renal de-fatting, maintaining fat over the tumour, laparoscopic ultrasonography to score the line of resection, en bloc hilar clamping, tumour excision with cold scissors, suture-repair of the collecting system and transected vessels, and sutured renorrhaphy using haemostatic agent and a Surgicel bolster.

The operating room set-up for RPN is shown in Fig. 1; inclining the operating table by 30–45° provides more room on the side of the robot to accommodate the vision cart, the laparoscopic boom and monitors, and for movement of personnel. Port placements for RPN and LPN are shown in Figs 2 and 3.

Figure 1.

The operating room system for left RPN.

Figure 2.

Port placement for (a) conventional right LPN; and (b) right RPN. (5 L, 5-mm liver retraction port; 12C, 12-mm camera port; 12S, 12-mm Satinsky port; 5A/12A, 5-mm/12-mm assistant ports; 8, robotic instrument ports. With the port-in-port technique, the 8-mm robotic ports are inserted through 12 mm laparoscopic ports.

Figure 3.

Port placement for (a) conventional left LPN; and (b) left RPN. Abbreviations as Fig. 2.

The operative steps of our RPN technique are detailed in Table 1. The robot is docked after the hilum has been prepared for clamping, and the kidney surface has been laparoscopically scored. After docking the robot, the table-side assistant clamps the hilum and the console surgeon excises the tumour robotically. Reconstruction of the kidney is completed robotically (similar to our LPN technique) and the hilum is unclamped. Additional sutures are placed as needed to ensure haemostasis. The robot is undocked and laparoscopic exit completed.

Table 1.  The steps of RPN
StepsCaveats/comments
  • *

    Bolstered renorrhaphy is performed after unclamping if an early unclamping technique is used.

  • †In the early unclamping technique, residual bleeding points in the PN bed are directly oversewn, thus providing definitive complete haemostasis. We think this will lead to a decrease in bleeding complications after surgery; long-term data on this are awaited.

Initial cystoscopic placement of ureteric catheterUse a glide wire to prevent ureteral trauma
Patient is placed in a 45° flank position for the transperitoneal approachAll pressure points must be padded and the patient secured to the table
Steps performed laparoscopically
 Creation of pneumoperitoneum and placement of all portsMaintain adequate distance and triangulation between ports
 Mobilization of the colonAvoid diathermy injury
 Mobilization of the kidneyThe kidney usually needs to be completely mobilized
 Preparation of the hilum for en bloc clampingIn case of multiple arteries, ensure all are clamped
 Strategic de-fatting of the kidneyMaintain fat over tumour. Maintain handles of fat on kidney to move and position the kidney
 Intraoperative ultrasonography and scoring line of resection on kidney surfaceReduces chances of a positive margin
Steps performed robotically
 Assistant applies laparoscopic Satinsky clampEnsure Satinsky clamp and port do not ‘fight’ the caudal robotic instrument and port
 Excision of the tumourAssistant should maintain bloodless field with expert but unobtrusive suctioning
 Running suture in the partial nephrectomy bedAimed to control transected blood vessels in the sinus fat and the collecting system. Due to absence of tactile feedback, there is a likelihood of suture breakage in the early part of a surgeon’s robotic experience. Note: when changing from conventional laparoscopy, the robotic needle drivers do not have a locking mechanism
 Retrograde injection to identify collecting system entry and suture repair as needed 
 Placing the bolster in the defect*Size the bolster to fit the defect. The bolster should not lie against the ureter, or it could cause ureteric obstruction after surgery
 Parenchymal renorrhaphy*Needs due care to avoid suture cut-through. Assistant unclamps the hilum
 Haemostasis is ensured. Additional sutures as needed
Steps after undocking the robot
 Specimen entrapment and retrievalFSA for margins. Patient remains under anaesthesia until FSA results are available for the entire specimen
 Drain placement
 Closure of port sites

Our LPN technique has undergone one significant recent refinement, in response to the remaining criticism of LPN, that the WIT is ≈10 min longer than for OPN, although this did not translate into any detectable detrimental effect on renal function [19]. To address this criticism we have developed an ‘early unclamping’ technique’ (Fig. 4) such that the hilum is unclamped after placing the first layer of sutures in the PN bed. Any residual bleeding points are then over-sutured under direct vision to secure complete haemostasis [20]. The pelvicalyceal system is repaired in a watertight manner. Bolstered renorrhaphy is then done as necessary in the perfused kidney. This technique was used in the last six patients treated with RPN. The matched LPN controls for these six patients were also treated with this technique to ensure comparability.

Figure 4.


Pictures from patient no. 9 (RPN): (a) preoperative CT showing the right lower-pole mass; (b) operative view of robotic excision of the mass with a rim of normal parenchyma; (c) operative view of the first running suture in the PN bed; and (d) operative view of the bolstered renorrhaphy.

RESULTS

Overall, the 12 patients treated with RPN were similar to the 12 matched controls treated with LPN in age, gender, BMI, ASA score, tumour size, side and location (Tables 2,3); Table 3 also shows the summary statistics for the pairs of six patients treated with the early unclamping technique. Perioperative outcomes were also comparable for operative blood loss, operation time, WIT and length of hospital stay. The overall mean WIT was 23 min for RPN and 22 min for LPN (P = 0.89). The mean estimated blood loss was 329 mL for RPN and 300 mL for LPN (P = 0.84). Renal functional outcomes, as measured by 3-month serum creatinine and estimated GFR (estimated by the Modification of Diet in Renal Disease equation) were comparable between the matched groups.

Table 2.  A matched-pair comparison of RPN and LPN; raw data
Type/age/sexBMI kg/m2ASA scoreSideLocationSize, mmWI time, minOR time, minEBL, mLBlood, unitsLOS, daysPatholComments/complications
  1. OR time, operating room time, EBL, estimated blood loss; LOS, length of hospital stay, Pathol, pathology, Onco, oncocytoma; CHF, congestive heart failure. *Asymptomatic perirenal haematoma on follow-up; †Robot malfunction while clamped (converted to LPN). Haemorrhage after surgery, transfused; ‡Perirenal haematoma; haematocrit decreased after surgery; §Atrial fibrillation after surgery; anticoagulated; ¶Pulmonary embolism, anticoagulated; bleeding after surgery required angioembolization; **Slow recovery of bowel function.

RPN/44/M281RLower2227360 900 6RCCConverted to LPN for PM
LPN/48/M331RLower2021270 100 5RCC
RPN/81/F282LMid2736240 200 3Onco
LPN/77/F272LMid3737195  50 4RCC
RPN/72/M292RLower3124225 150 3RCC
LPN/68/M262RLower3521360 200 5RCC*
RPN/48/F252LMid1424180 100 2Cyst
LPN/50/F202LMid2042120 150 3AML
RPN/53/M292RLower1616130 1503 4RCC
LPN/56/M322RLower2232360 250 4RCC
RPN/55/F473RUpper3819240 400 4RCC
LPN/46/F463RUpper3032250 2001 5RCC
RPN/80/M252RUpper2320330 350 8RCCCHF
LPN/72/M302RUpper331532015001 4RCC
RPN/78/M242RLower2030180 200 3RCC
LPN/80/M292RLower2510270 200 7Onco§
RPN/73/M252RLower2213240  50410RCC
LPN/71/M252RLower2210240 200 3RCC
RPN/77/F353LMid2526180  50 3RCC
LPN/72/F333LMid3419180 3001 4RCCSlow decline in haematocrit
RPN/52/M251LMid3323300 400 6RCC**
LPN/43/M301LMid4113270 150  4RCC
RPN/56/M281LLower21153001000 4RCC
LPN/46/M261LLower3116240 300 5RCC
Table 3.  Summary statistics of the matched-pair analysis of 12 RPN with 12 LPN
VariablesAll 12Six with early unclamping
RPNLPNPRPNLPNP
Gender, M/F  8/4  8/41  5/1  5/11
Side, R/L  7/5  7/51  3/3  3/31
Location, lower/mid/upper  6/4/2  6/4/21  3/2/1  3/2/11
Mean (sd):
Age, years 64 (13.8) 61 (13.8)0.37 69 (12.1) 60 (16.4)0.12
BMI, kg/m2 29 (6.4) 30 (6.4)0.76 27 (4.1) 31.5 (7.6)0.26
ASA score*  2 (1–3)  2 (1–3)1  2 (1–3)  2 (1–3)1
Tumour size, mm 24 (6.9) 29 (7.1)0.06 24 (4.7) 30.5 (6.7)0.01
Estimated blood loss, mL329 (315)300 (384)0.84342 (354.1)442 (521.9)0.70
 Median (interquartile range)200 (112–400)200 (150–287) 275 (50–550)250 (187–600) 
WIT, min 23 (6.6) 22 (10.8)0.89 21 (6.5) 14 (3.5)0.05
Operating room time, min242 (69.2)256 (70.6)0.60255 (65.0)253 (46.3)0.93
Hospital stay, days  4.7 (2.4)  4.4 (1.1)0.77  5.7 (2.8)  4.5 (1.4)0.50
Serum creatinine, mg/dL
 Before surgery  0.97 (0.2)  1 (0.2)0.73  1 (0.14)  1 (0.2)1
 3 months after  1.12 (0.2)  1.15 (0.2)0.74  1.1 (0.18)  1.2 (0.26)0.56
Estimated GFR, mL/min
 Before surgery 88 (22.2) 85 (21.9)0.77 74 (15.9) 75 (17.7)0.82
 3 months after 75 (22.6) 72 (17.0)0.71 68 (23.6) 63 (16.1)0.69
Follow-up, months  7.4 (5.2)  8.5 (5.6)0.90   

Two patients in the RPN group were converted to standard LPN, one for a positive focal margin on frozen-section analysis (FSA), and the other for a robotic camera malfunction. The positive margin on FSA occurred in the first patient treated with RPN. In this patient, RPN had been completed, the robot had been undocked, and the excised tumour sent for FSA as usual. When the margin was reported focally positive, we made an additional resection of the PN bed using standard LPN to achieve negative margins. The WIT for this additional resection and reconstruction was 13 min. In the fifth patient treated with RPN, the robotic light source malfunctioned 3 min after the hilum was clamped. The robot was quickly undocked and the procedure completed laparoscopically with no further incident. This experience led us to develop a ‘port-in-port’ technique; in this technical modification, the right-hand and left-hand working ports placed for the initial laparoscopic component of the RPN are actually standard 12-mm laparoscopic ports. When it is time to dock the robot, the robotic 8-mm ports are inserted inside these 12-mm ports and the robotic operation is continued. This arrangement provides added confidence that in event of a robotic malfunction, we can undock expeditiously and proceed with standard LPN in the usual fashion, without wasting time in inserting new ports or working through 8-mm robotic ports that do not allow the passage of CT-1 or CT-X needles.

DISCUSSION

The daVinci system (Intuitive Surgical, Sunnyvale, CA, USA) has allowed laparoscopically naive urologists to perform laparoscopic prostatectomy, by providing them the more familiar three-dimensional operative view, and the ‘easier’ wristed instruments and needle drivers. Before the da Vinci system entered mainstream urology, minimally invasive prostatectomy was performed laparoscopically by a few surgeons who possessed the requisite technical skills. The robot changed this situation, enabling more urologists to conduct a minimally invasive prostatectomy, which they were earlier unable to do. Patient demand and marketing pressures made robotic prostatectomy hugely popular in the USA. Interestingly, in Europe, the system is not as prevalent, as many more urologists in continental Europe are trained in standard laparoscopy, and the healthcare system and reimbursements are different.

PN is an advanced procedure; open surgery involves a large muscle-cutting, often rib-resecting, flank incision. Laparoscopically, rapid excision and complex suturing is required, often at difficult angles. Dissemination of LPN has been slow, primarily because of the perceived technical difficulty involved. Can robotics be as successful for LPN as it has been for radical prostatectomy?

Since 2004, four centres have reported their experience with RPN [13–17] (Table 4). The published global experience to date is <50 patients, compared to >2000 patients for LPN. Although the feasibility of RPN is confirmed, no clear advantage of RPN over LPN has been apparent. However, the important concept is that for RPN to be successful there need be no advantage over LPN; that the robot might enable more urologists to perform LPN is enough to make RPN a successful procedure, analogous to the success enjoyed by robotic prostatectomy.

Table 4.  Previously published studies of RPN compared with the present report
VariableStudy
[13][14]*[15]*[16][17]Present
  1. UP, upper pole; MP, mid pole; LP, lower pole; NS, not stated; trans, transperitoneal; retro, retroperitoneal; HAL, hand-assisted laparoscopy; PE, pulmonary embolism; LOS, length of hospital stay; EBL, estimated blood loss; OR, operating room; NR, no recurrence. *[14,15] pertain to the same dataset from the same centre. †Not including 20-mim robot set-up time. ‡Some smaller lesions were excised without hilar clamping after the index tumour was excised with the hilum clamped.

Type of studyCase seriesCase series (technique report)Case-controlCase seriesCase seriesMatched pair
No. RPN131210108 (14 tumours)12
No. LPN1012
Mean (range) tumour size, cm3.5 (2–6)1.822.3 (1–3.5)2.4 (0.8–6.4)2.4 (1.4–3.8)
Side, R/L7/6NSNS6/43/57/5
Tumour locationUP 3; MP 5; LP 5NSUP 3; MP 3; LP 4UP 3; LP 3; NS 4Hilar 5; UP+MP 1 UP+ LP 1; UP 1UP 2; MP 4 LP 6
TechniquePure robotic 11 trans, 2 retroRobot-assisted All transRobot assisted All transPure robotic All transPure robotic All transRobot-assisted All trans
Hilar clamp methodBulldogsBulldogsBulldogsBulldogsBulldogsSatinsky
Mean (range)
OR time, min215 (130–262)265 (NS)279 (NS)155 (120–185)192 (165–214)242 (130–360)
EBL, mL170 (50–300)240 (NS)24092230 (100–450)329 (50–1000)
WIT, min(five) 22 (15–29)26 (NS)26.421 (18–27)31 (24–45)23 (13–36)
Cold IT(eight) 33 (18–43)     
+ve margin on final pathology1NSNegativeNegativeNegativeNegative
Mean LOS, days4.3 (2–7)2.72.63.5 (1–21) median 1.52.6 (2–3)4.7 (2–10)
No. RCC10NS8879
Follow-up, months2–11, NRNSNS15 (6–28) NR3 NR7.4 NR
CommentsIleus, 1Conversion to HAL 1, SL 1, open 1Conversion to HAL 1, open 1 retention 1Urine leak 1 Re-exploration for bleed 1Converted to LPN 2; PE and angio- embolization 1

With this in mind we undertook the present study to compare RPN with conventional LPN, the main purpose being not to determine whether RPN is better than LPN, but rather to evaluate whether RPN can be done in a technically comparable manner to LPN. For this, our large previous experience with LPN in >800 cases, and the resultant understanding of peri-renal and intrarenal surgical anatomy and familiarity with technical nuances of LPN, were brought to bear.

Rigorous case-by-case matching was used to ensure that controls were as closely comparable as possible. Overall, although there was no obvious advantage of RPN over LPN in our hands, RPN was no statistically worse than LPN either. However, when patients undergoing our current ‘early unclamping’ technique were compared, the WIT was substantially shorter with LPN than RPN.

Subjectively, the RPN procedure appeared to be more difficult for the entire operating team than was LPN. This remained true from the first RPN case to the last. This might be because our operating team is experienced in conventional LPN, whereas it is still gaining experience with RPN.

RPN is technically feasible and safe; it requires training and experience, although how much is difficult to quantify, and whether the requirements would differ for a surgeon who has more experience with robotic prostatectomy is unknown.

There are some caveats; RPN is a technically challenging procedure, primarily due to the time constraints posed by hilar clamping. Although a surgeon has the liberty to take 6 h for his/her first robotic prostatectomy, such a luxury is not available while suturing during RPN. Thus before embarking on the first RPN, the surgeon should be well versed with handling and troubleshooting the robotic interface. Familiarity with the laparoscopic environment and three-dimensional renal anatomy is important to plan and execute tumour excision and skilled suture reconstruction. A robotic malfunction during robotic prostatectomy allows a surgeon some time to fix the problem. This is not possible if the hilum is clamped during RPN. Therefore, ideally, the surgeon who embarks upon RPN should have the skills and confidence to be able to undock the robot and expeditiously proceed laparoscopically if such a malfunction occurs at a critical juncture, as occurred in one of our cases. Finally, it is important to note that cases for RPN in the present series were selected carefully, and generally were technically less challenging ‘easier’ cases than what is the norm during LPN at our centre.

Currently the robot confers no tangible advantage over LPN. However, RPN is safe and likely to be used for minimally invasive PN. RPN could potentially enable more surgeons to perform LPN, thereby allowing wider dissemination of MINSS. However, it is incumbent upon us to ensure that this dissemination is responsible, and done carefully so that the patient benefits unequivocally.

CONFLICT OF INTEREST

None declared.

Ancillary