Surgical outcomes of robot-assisted partial nephrectomy
Brian M. Benway, Division of Urologic Surgery, Washington University School of Medicine, Saint Louis, MO, USA. e-mail: firstname.lastname@example.org
What's known on the subject? and What does the study add?
Robot-assisted partial nephrectomy has been established as a safe and effective approach for nephron-sparing surgery. Early oncological and functional outcomes appear to be on a par with other established approaches.
This review provides an up-to-date overview of contemporary experience with robot-assisted partial nephrectomy. We also discuss techniques which may help to optimize outcomes of the procedure.
- • Robot-assisted partial nephrectomy (RAPN) has emerged as a viable technique for nephron-sparing surgery.
- • In this article, we review the outcomes of RAPN in its current iteration, with attention to operative, oncological, and functional outcomes. In addition, we highlight techniques that may serve to enhance outcomes.
(robot-assisted) (laparoscopic) partial nephrectomy
(laparoscopic) radical nephrectomy
warm ischaemic time
positive surgical margin
RCC is often found incidentally on abdominal imaging performed for typically unrelated purposes. The result has been a shift in the diagnosis of renal malignancy towards smaller masses that are amenable to nephron-sparing management [1–4].
Partial nephrectomy (PN) was popularized in the 1980s as an alternative to radical nephrectomy (RN). In the ensuing years, long-term analyses have found that in addition to providing equivalent oncological control to radical extirpation [5–7], the preservation of healthy renal tissue is associated with increased overall survival and a decrease in the development of significant comorbid disease [8,9].
In 1993, McDougall et al.  and Winfield et al.  described the first cases of laparoscopic PN (LPN). While capable of providing acceptable outcomes in the hands of an experienced minimally-invasive surgeon, LPN was found to be a challenging procedure, with a long learning curve [12,13]. Interestingly, the use of PN actually decreased in the first part of the millennium, while use of laparoscopic RN (LRN) increased proportionally . In fact, at last evaluation, nephron-sparing techniques were found to be used in <50% of patients with tumours of <2 cm and <20% of patients with lesions of 2–4 cm .
In 2006, Gettman et al.  first described robot-assisted PN (RAPN); however, early studies evaluating RAPN failed to find tangible advantages to a robot-assisted approach [17–19].
However, over the past 5 years, several refinements of technique have been introduced, and subsequently, RAPN has become a reasonable alternative to LPN and open nephron-sparing techniques. In this review, we will focus upon outcomes associated with RAPN in the modern era.
OPERATIVE PARAMETERS, PERIOPERATIVE COMPLICATIONS, AND FUNCTIONAL OUTCOMES
Over the last 5 years, RAPN has proven to be a safe and efficacious procedure, with operative parameters on a par with laparoscopic and open approaches. A summary of outcomes of large contemporary series can be found in Table 1[20–35].
Table 1. Overview of contemporary RAPN series with ≥20 patients
|Rogers et al., 2008 ||148||2.8||197||27.8||183||1.9||9 (6)||2 (1)||6 (4)||NSM: 7.2 (2–54), NR|
|PSM: 18 (12–23), NR|
|Bhayani and Das, 2008 ||35||2.8||142||21.0||133||2.5||6 (17)||1 (3)||None||NR|
|Wang and Bhayani, 2009 ||40||2.5||140||19||136||2.5||6 (15)||1 (3)||1 (3)||≤12, NR|
|Ho et al., 2009 ||20||3.5||82.8||21.7||189||4.8||None||None||None||>12, NR|
|Benway et al., 2009a ||50||2.7||145.3||17.8||140.3||2.5||5 (10)||2 (4)||1 (2)||12, NR|
|Benway et al., 2009b ||129||2.9||189||19.7||155||2.4||10 (8)||2 (2)||5 (4)||≤12, NR|
|Jeong et al., 2009 ||31||3.411||169.9||20.9||198.3||5.2||1 (3)||None||1 (3)||12, 1 recurrence|
|Scoll et al. 2010 ||100||2.8||206||25.5||127||3.2||13 (13)||2 (2)||5 (5.7)||12.7, NR|
|Benway et al., 2010 ||183||2.9||210||23.9||132||Not reported||18 (9.8)||2 (1)||7 (3.8)||≤26, NR|
|Haber et al., 2010 ||75||2.75||200||18.2||323||4.2||12 (16)||3||None||Not reported|
|Patel et al., 2010 ||71||2.1–5||238–275||20–25||100||2||9 (12.6)||1 (1.4)||3 (4.2)||12, NR|
|Gong et al., 2010 ||29||3.0||197||25.0||220||2.5||None||Not reported||None||15, NR|
|Petros et al. 2010 ||95||2.3–2.5||246–250||16–21||100–150||2||22 (23)||None||Not reported||Not reported|
|Williams et al. 2011 ||27||2.47||233||18.5||179.6||2.51||5 (18.5)||None||1 (4)||Not reported|
|Lorenzo et al., 2011 ||65||Not reported||171||Not reported||243.2||4.6||1 (1.5)||1 (1.5)||6 (9.2)||13, 5 recurrences|
|Naeem et al., 2011 ||97||2.3–2.5||242.5–265||22.5–26.5||100–150||2||8 (8.2)||1 (1)||2 (2)||12, NR|
|Dulabon et al., 2011 ||446||2.88–3.46||187.4–194.5||19.6–26.3||208.2–262.2||2.87–2.94||23 (5.2)||10 (2.2)||7 (1.6)||21–45, 1 recurrence|
Operative durations in contemporary series range from 82 to 275 min, and are typically longer for larger renal tumours , complex and hilar renal masses [21–23], and for those patients with prior abdominal surgery . In addition, RAPN performed through a single-site approach is associated with operative durations as low as 170 min , but are often >200 min .
Warm ischaemic time (WIT) has been a primary focus of studies pertaining to minimally-invasive nephron-sparing surgery (NSS), as reduction of WIT is often seen as advantageous in maximizing preservation of renal functional reserve. As with LPN, tumour excision and renal reconstruction are often carried out under warm ischaemia. While there is much debate surrounding the maximum acceptable WIT, it has been suggested that in patients with normal preoperative renal function, WIT of up to 25–30 min is well-tolerated, with no significant change in postoperative renal function [38,39].
While longer WITs are not uncommon during initial experiences, it is perhaps then encouraging that in all contemporary series of large (>20) patient cohorts, the mean WIT is consistently below the 30-minute watermark, with WITs ranging from 16 to 27.8 min [20–35,40]. While WITs tend to increase with increasing tumour complexity , renal masses >4 cm , and hilar renal masses , WITs of <30 min can be consistently achieved through a robot-assisted approach.
Even in those patients for whom tumour complexity renders short WIT difficult to achieve, mounting evidence suggests that minimally-invasive NSS nevertheless confers distinct advantages over alternatives, including RN. In one recent analysis, Lane et al.  reported that even for WITs of up to 144 min, patients who underwent PN were far less likely to develop postoperative renal insufficiency than comparable patients who underwent RN.
Indeed, WIT may have less influence over postoperative functional outcomes than previously thought. In a large series of 660 patients undergoing PN, WIT duration was found to have no statistically significant influence upon postoperative renal function. Rather, it was the patient's preoperative renal function, coupled with the amount of parenchyma spared that appeared to influence functional outcomes, suggesting that non-modifiable patient parameters may be the main determining factors in functional outcomes after PN .
As mentioned above, tumour complexity can influence certain operative parameters. Dulabon et al.  conducted a large retrospective analysis comparing RAPN for hilar vs non-hilar masses, finding that hilar masses were associated with a significantly increased WIT (26.3 min) vs non-hilar masses (19.6 min, P < 0.001); however, other parameters, such as overall operative duration and intraoperative blood loss were similar. Interestingly, this study also found fewer conversions, complications, and positive margins in the hilar tumour group, but this probably represents surgeon experience, as more challenging cases were tackled further along the learning curve.
These findings corroborate an earlier analysis by Benway et al. , who found that tumour complexity had a significant influence upon WIT (15.3 min for simple tumours vs 25.9 min for complex masses, P = 0.01), although the operative durations between the two groups were equivalent. These outcomes compare favourably to LPN performed by the same surgeons, for whom tumour complexity increased both overall operative duration and WIT, with significantly longer WIT for LPN than RAPN in both cohorts.
Complication rates across most mature RAPN series appear to be comparable with those of LPN. In the two largest series of pure LPN to date, Gill et al. [12,43] found an overall complication rate of 18.6% for LPN across all patients, with a complication rate decreasing from 22.1% for initial procedures to 8.5% for the procedures after the learning curve had been reached. As can be seen in the overview of mature series in Table 1, complication rates for RAPN are similar to that of LPN, ranging from zero to 18.5% for typical RAPN series. Petros et al.  report the only complication rate >20%, but their study is notable for focusing upon patients who had undergone previous abdominal surgery, and it is therefore unsurprising that their complication rates would be slightly higher than other series.
One possible explanation for the relatively early achievement of comparable operative outcomes is the foreshortened learning curve of RAPN compared with LPN. In a previous study, Link et al.  evaluated the learning curve of LPN, finding that mastery of critical portions of the case could not be attained even after 200 procedures. Indeed, in their experience of 800 patients, Gill et al.  found that the learning curve for WIT was not met until after 565 procedures. By comparison, the learning curve for RAPN for portions of the procedure performed under warm ischaemia has been found in multiple studies to be ≤30 procedures [26,44,45].
ONCOLOGICAL OUTCOMES AND CANCER CONTROL
Numerous studies have shown the oncological efficacy of minimally-invasive NSS. Indeed, despite the added insult of warm vs cold ischaemia and the relatively rigid instrumentation, several studies have reported that LPN is capable of providing equivalent oncological and functional outcomes to open PN [43,46–48]. Likewise, several recent studies have found that RAPN provides oncological outcomes on par with LPN [21,29,31,34,35,40–53].
As can be seen in Table 1, early and intermediate outcomes show excellent oncological control. In fact, across >1600 patients detailed in modern large series, only seven recurrences have been reported to date, a rate of <1%. While certainly encouraging, it is important to note that as RAPN is a relatively new procedure, long-term data is presently lacking.
Across the studies covered in this contemporary review, there have been 45 (2.7%) reported positive surgical margins (PSMs), some of which have accounted for the recurrences reported above. However, it is important to note that a PSM does not always indicate residual disease. In a large series of >770 patients undergoing PN, Kwon et al.  found that the incidence of recurrent disease was equivalent between those with PSMs and those with negative surgical margins, and that less aggressive tumours were unlikely to recur, regardless of the presence of a PSM. Indeed, more recent analyses have found recurrences after PSM to be very rare [55,56].
Naturally, the lack of strong correlation between PSM and recurrence can be a confounding issue when determining the need for re-resection or completion nephrectomy. In a recent analysis of 29 patients who underwent re-resection or completion nephrectomy for intraoperative PSM, none of the patients who underwent outright completion nephrectomy were found to harbour residual malignancy. Furthermore, only two patients were discovered to have a PSM upon re-resection; these patients were found to harbour more extensive malignancy upon completion nephrectomy .
Taken together, these results suggest that most patients with a PSM will not develop disease recurrence. Should suspicion persist, re-resection might be considered if feasible, but completion nephrectomy may be overtreatment in most instances.
OPTIMIZING OUTCOMES OF RAPN
Several refinements in approach and technique over the past 5 years have helped to transform RAPN into a first-line intervention for organ-confined renal tumours. The most important of these considerations are discussed below.
For the inexperienced robotic renal surgeon, proper patient selection for the first few cases is critical. The lack of haptic feedback and the reliance upon the bedside assistant can present challenges unique to RAPN. Ideal candidates for initial RAPN procedures include patients with predominantly exophytic T1a lesions, uncomplicated vascular anatomy, and a normal contralateral kidney. Female patients are preferred due to the relative paucity of perirenal fat. Posterior upper pole masses, hilar masses, and deeply endophytic tumours should be avoided during the initial experience, as should patients with a solitary kidney or severely impaired baseline renal function. Beginning with RARN may help the surgeon develop familiarity with hilar dissection under robotic control.
CHOOSING THE BEST APPROACH
The transperitoneal approach appears to be the most commonly used approach for RAPN, probably owing to its predominance in pure laparoscopic renal surgery. While there are multiple small variances in port placement between institutions, there are two main trocar configurations for transperitoneal RAPN, each with certain advantages and disadvantages.
The most commonly used camera and trocar configuration places the camera in a medial position, near the umbilicus. A 30° downward-angled lens is used. Trocars for the robotic arm are placed just cephalad of the anterior superior iliac spine, and a few centimetres below the costal margin in the mid-axillary line. One assistant port is then placed at the midline in one of two locations, depending upon tumour location. If a fourth arm is to be used, the trocar is placed laterally, triangulated between the two other trocars [26,57–61].
The advantages of the medial camera approach are a wide viewing angle and a perspective that is familiar to surgeons with prior laparoscopic experience. This angle allows for simple visualization of the bowel during mobilization, and allows for the camera to be easily panned allowing instruments to be passed into the field under direct vision. Furthermore, the placement of the assistant port generally offers unhindered access for the assistant, with minimal interference from the robotic arms. One potential disadvantage of this approach is the distance to the kidney, as this often exceeds the reach of the camera arm's excursion. However, this issue rarely presents difficulty, and is largely offset by the digital zoom offered by current generation robotic systems .
The alternative approach uses a lateral camera position, and a 30° upward-angled lens. Trocars for the robotic arm are then placed in a similar, but more medially displaced fashion as the medial camera approach. The fourth arm is placed a few centimetres lateral to the umbilicus. Two assistant ports are typically placed [59,62,63].
Advantages of the lateral approach include the comparatively shorter distance to the kidney, which allows for closer visualization. In addition, this approach may be better suited for posterior tumours, which can often prove quite challenging using a medial-camera approach. Disadvantages include the inability to zoom out to a wide angle, which can increase the technical challenge of bowel mobilization, and can hinder the ability to pass instruments under direct vision. In addition, the assistant must often work on either side of the fourth arm .
Retroperitoneal approaches for RAPN are well-suited for posterior masses, or for those patients with a history of previous abdominal surgery that may compromise a transperitoneal approach. While safe and effective in experienced hands, the retroperitoneal approach nevertheless may be challenging due to a relatively confined workspace and a fewer landmarks to help orient the surgeon .
PREOPERATIVE AND INTRAOPERATIVE IMAGING
Whenever possible, contrast-enhanced CT should be obtained preoperatively to define the vascular anatomy. Renal hilar anatomy varies between patients, and multiple arteries and veins are not an uncommon finding. Advance knowledge of these variations in vascular anatomy can greatly aid hilar dissection and prevent inadvertent vascular injury to an anomalous branch.
To ensure satisfactory oncological control, most surgeons advocate the use of intraoperative ultrasonography (US) to aid in the delineation of the extent of the tumour. Newer robotic platforms include TilePro software integration, which allows for real-time picture-in-picture display of imaging in the surgeon's viewfinder. This allows the surgeon to directly correlate the imaging with the view of the surgical field, simplifying the process of mapping out the extent of dissection. In addition, the TilePro software allows for display of other imaging, including CT from a connected computer [65,66]. Our current approach uses robot-controlled US to allow complete surgeon control of intraoperative imaging.
TUMOUR EXCISION AND RENAL RECONSTRUCTION
After identification of the renal lesion and choosing between clamping the main artery, a branch thereof [67–69], or performing the surgery ‘off clamp’[70–73], the tumour is then resected in entirety. Imaging guides the tumour resection, and deep structures such as the renal sinus or collecting system may be entered. After complete excision, the kidney is reconstructed.
For closure of the renal defect, sliding-clip renorrhaphy has been described as a preferable alternative to traditional tied-suture closures. This renorrhaphy method relies upon the use of Weck Hem-o-Lok clips placed on either side of the defect, which are then slid into place by the surgeon to place tension upon the repair. The Hem-o-Lok clip is often then locked in place with a LapraTy clip, which prevents backsliding of the clips [19,26,58,74,75]. By simplifying the closure, sliding-clip renorrhaphy has been clearly shown to significantly reduce WITs by up to 13 min compared with traditional tied-suture renorrhaphy .
RAPN is a safe and efficacious technique for minimally-invasive NSS with a foreshortened learning curve that provides a comparatively low barrier of entry over pure laparoscopic techniques. Contemporary series clearly show that RAPN is capable of providing excellent functional and oncological outcomes, with operative parameters and perioperative complication rates equivalent to, and in some instances superior to, both open and laparoscopic approaches. Careful patient selection and meticulous preoperative planning, in addition to real-time intraoperative imaging and techniques for minimizing ischaemic insult, help to optimize the outcomes of RAPN.
CONFLICT OF INTEREST
Brian M. Benway is a Consultant to Vikins Systems.