A comparison between laparoscopic and robotic-assisted laparoscopic pyeloplasty in patients with ureteropelvic junction obstruction

Methods


Background
In 1891, Kuster described the rst successful pyeloplasty which has been considered the gold-standard treatment for ureteropelvic junction obstruction (UPJO), with success rate exceeding 90% (90-100%) [1 − 3]. Later within these two to three decades, a procedure of minimally invasive surgery (MIS) has been developed and gradually famous and become an accepted treatment option with varieties procedures including endoscopic pyelotomy, laparoscopic pyeloplasty (LP) and robotic-assisted laparoscopic pyeloplasty (RALP).
Although LP and RALP present success rates at a level comparable to those of open surgery, these two procedures has provided the advantages of lower morbidity, less postoperative pain, faster recovery time and shorter hospital stays [4 − 7]. Schuessler et al. [8] were the rst one to do laparoscopic pyeloplasty in 1993. Since then, the procedure have published its popularity. However, anastomosis of ureteropelvic procedure is still considered as a hard and challenged for LP. The procedure demands advanced laparoscopic skill with a present of steep learning curve. These are limitations for a novice surgeon with low laparoscopic surgery skill.
With high technology development of robotic-assisted surgery, limitations occurred in LP surgery procedure have been gradually demolished. The procedure becomes less complex operative task, especially, reconstructive procedures for anastomosis suturing. These circumstances contribute towards the ongoing goodwill of RALP. However, robotic system exhibits limitations on lacking of tactile feedback sensation, wasting of docking time for setting up an assisting surgical robot instruments and highly supportive disbursement in several countries [9].
Up to date, there's still no con rmation that RALP has signi cant impact on outcomes over LP. To explore this issue, this retrospective study aimed to compare outcomes between LP and RALP in patients with UPJO by our single center.

Design and setting
The study protocol and approval to conduct this study was obtained from the local institutional research and ethics Board. The medical records of all patients who underwent LP or RALP due to UPJO between March 2008 and May 2019 were retrospectively reviewed. All patients were con rmed diagnosis of UPJO by renal ultrasonography (US), intravenous pyelogram (IVP) (optional), computerized tomography (CT) scan and a diuretic renogram. The indications for surgery are renal pain, urinary tract infection (UTI), stone formation, decrease renal function (split renal function < 40%), obstructive pattern on diuretic renogram, progressive hydronephrosis on serial US examination and progressive loss of renal function on serial radionuclide renograms. The LP has been previously performed in 2008 through 2012. Since 2013 when the da Vinci ® Si robotic system (Intuitive Surgical, Sunnyvale, CA) arrived in our hospital, all patients has been undergoing RALP.
The data collected from the chart review included patient demographics, preoperative evaluation, perioperative data (including total operative time, robotic docking time, and robotic console time), and postoperative results. Operative time was de ned from the time of skin incision to skin closure, including docking time. Perioperative and postoperative complications were classi ed based on Clavien and Dindo classi cation [10]. The degree of hydronephrosis was grade of 0 to 3 according to Ellenbogen's grading system [11].
The primary outcome of our study was the success rate of surgery. This was de ned as relief of symptoms as well as improve obstruction on a diuretic renogram at 6 to 12 months after surgery. The secondary outcome included operative time, length of hospitalization (LOH), estimated blood loss (EBL) and complications.

Surgical procedure
The Anderson-Hynes dismembered technique was performed in all patients for both LP and RALP group.
The LP was performed retroperitoneally, whereas the RALP was an intraperitoneal approach. All patients were positioned in the standard ank position.

Retroperitoneal laparoscopic pyeloplasty
Our technique for LP has been described in the previous study [12]. A camera port was made in the middle axillary line at the midpoint between the 12th rib and the iliac crest. A balloon dissector was used to create retroperitoneal space, a 12-mm trocar was placed. One working port was placed 7 cm dorsal to the camera port and two ports were placed in the anterior axillary line at the level of the iliac crest and in the subcostal region. Following exposure of the psoas muscle and the Gerota's fascia, the ureter was identi ed and mobilized to the ureteropelvic junction (UPJ). The stay suture was placed passed to the renal pelvis and pulled out through the abdominal wall. The ureter was spatulated laterally and the stenotic segment of the UPJ was resected. A 6F ureteral stent was inserted into the ureter by antegrade fashion. The anastomosis was performed with a running suture using a 4-0 polydioxanone suture. The anterior anastomosis was initially completed, followed by the posterior anastomosis with the same suture. The crossing vessels were preserved by anterior transposition of ureter and pyelo-reduction was performed if necessary. If renal calculi were present, the stones were removed before anastomosis by laparoscopic graspers or by using exible nephroscope with a stone basket through the 12-mm port.

Robotic-assisted laparoscopic pyeloplasty
The transperitoneal approach was our preferred in RALP. The position of the ports as shown in Fig. 1. A 12-mm camera port was made 6 cm lateral to the umbilicus by open Hasson technique. The 8-mm robotic arm ports were placed laterally at a distance of 8 cm from the camera port. The rst assistant port was set at a distance of 6 cm from the midpoint between the caudal robot's arm port and the camera port. The da Vinci ® robotic system was placed on the dorsal side of the patient.
The operation started with colon mobilization, allowing the colon to fall medially. Gerota's fascia was exposed. The ureter was identi ed and dissected up to UPJ. The rest of the surgical steps were almost the same as LP, except RALP did not need to stabilize the renal pelvis by stay sutures. Because of freedom of robotic arms, it was easy to perform anastomosis without any xation.

Follow up
The ureteral stent was removed at 4 weeks after surgery. Follow up was scheduled for evaluation of clinical, urine examination and the renal US. A diuretic renogram was also performed at 6 to 12 months postoperatively.

Statistical analysis
We used the Student T-test for comparison between two groups that showed equal variance. On the other hand, the Mann-Whitney test was used between two groups showing non-normal distribution or non-uniform variance. In addition, in comparison of the ratio between the two groups, a chi-square test was used. For each test result, a corresponding two-sided p-value of <0.05 was considered statistically signi cant. All analyses were performed using SPSS software (IBM, Statistical Package for the Social Sciences ver 23, Chicago, IL USA).

Results
A total of 25 patients were included in this study. Thirteen and 12 patients underwent LP and RALP, respectively. Patient demographics and characteristics data are shown in Table 1. There were signi cantly higher BMI in LP group than RALP group (P = 0.012). However, the gender, age, side, clinical presentation, and history of previous surgery were similar in both groups. The crossing vessels were encountered in 5 (39%) patients in LP group and 4 (33%) patients in RALP group and they could be preserved in all cases. Two (15%) patients in LP group had concomitant stones compared with 2 (17%) patients in RALP group. The median operative time for RALP, including docking time, was signi cantly shorter than LP (296 minutes vs. 199 minutes in LP and RALP, respectively) (P = 0.001), while the median robotic docking time was 6 minutes (range 2 to 17) ( Table 2). There was no signi cant difference in the median EBL between LP and RALP (5 ml vs 0 ml, respectively) (P = 0.168). However, 2 cases in LP group had signi cant blood loss (305, and 600 ml), but all of them did not require a blood transfusion. Regarding complications, no intraoperative complications were observed in both groups. While in postoperative period, one patient in the RALP group was Clavien-Dindo level IIIa. This patient had ureteral stent obstruction by a blood clot, which required to change stent by cystoscopy under local anesthesia. The median time for drain removal in LP group was 3 days (range 2 to 4), which is signi cantly more than 2 days (range 1 to 4) in RALP group (P = 0.029). The median length of hospital stay after surgery was similar in both groups were equivalent at 8 days (P = 0.677). The success rate, de ned as resolution of symptoms and improvement of obstruction on a diuretic renogram, were similar at 100% for both groups. No open conversions were required for all patients. The median follow-up was 24 months.

Discussion
At the present, LP and RALP surgery have become alternative procedures with a minimally invasive surgery for the treatment of UPJO. The highly popular procedures displayed treatment effectiveness at a high level comparative to that of open pyeloplasty which once was the gold-standard surgery in the past [13]. Furthermore, these procedures provided the advantages of lower morbidity, less postoperative pain, and shorter hospital stays. Nevertheless, it was found that LP demanded high intracorporeal suturing skill and longer operative time comparative to open surgery [14]. After the robotic system has been introduced to operative treatment, various limitations on laparoscopic surgery could be demolished including decreased learning curve, reduced fatigue of a surgeon during surgery operation and increased effectiveness of operative treatment. With 3 dimensional visualization (3D) including freedom movement of robotic wrist, stable view, tremor ltering, and motion scaling, RALP procedure has gained higher popularity among surgeons [15,16].
Braga has conducted a meta-analysis to compare LP vs. RALP treatments from 8 research studies which showed success rate exceeding 90%, and 5 of which exhibited success rate at 100% for both LP and RALP. It was found that success rates of RALP were higher than those of LP (100% versus 97% and 99% versus 97%) in 2 studies. In contrast, there was only one study reported that success rate of LP was greater than that of RALP (100% versus 97%). However, the difference showed no statistical signi cance [17]. Moreover, our study found that success rates of both LP and RALP were equivalent at 100% when evaluation was performed from relief of symptoms and improve obstruction on a diuretic renogram. Furthermore, there was no open conversion discovered in both groups. Additionally, our study has disclosed that median operative time of LP was 296 (185-498) minutes, whilst that of RALP was 199 (173-296) minutes. The results suggested that RALP employed shorter total operative time comparative to that of LP with statistical difference. Light et al. [18] has performed a meta-analysis study to compare the operative time of LP and RALP in both children and adults from 14 research studies. The results indicated that there were 4 studies showing comparatively identical operative time in both LP and RALP. Whilst, there were 8 studies exhibiting shorter operative time in RALP compared to that of LP with statistical signi cance. In contrast, there were 3 studies showing that LP statistically signi cantly display shorter operative time comparative to RALP. According to the results from meta-analysis studies performed by Light et al., it can be concluded that RALP employed operative time shorter than LP for 27 minutes. Interestingly, it was believed that assisted robotic surgery involved shorter anastomosis suturing comparative to that of laparoscopic surgery [18].
However, Link et al. [19] has reported that LP displayed operative time shorter than that of RALP with statistical signi cance. Furthermore, RALP surgery could expedite anastomosis suturing procedure, though, some operative steps, such as colon re ection, consumed more time comparative to laparoscopic surgery since the robotic arms have been designed to operate with more effectiveness for precise than gross movements. Additionally, the robotic system consumed more time for robot docking and undocking [20]. Moreover, especial novice surgery center with low experience in robot-assisted surgery would consume more time for setting up procedure. Therefore, the operative time not only depended upon surgeons, but also experience and masterfulness of the entire robotic team [18,21].
Lucas et al. [22] has reported a study comparing the surgery results by LP and RALP, performed in 274 and 465 patients, respectively. No statistically signi cant difference has been found in intraoperative complications, postoperative complications and anastomotic leak. In our study, intraoperative complications were not observed in both groups. However, it has been found that there was intraoperative signi cant blood loss among 2 patients operated by LP (305 and 600 ml, respectively), but blood transfusion was not required. Furthermore, one case in this group was previous performed balloon endopyelotomy with no success, therefore, it was di cult during LP surgery as a result of adhesion leading to renal parenchymal tear. Moreover, bleeding from inadequate renorrhaphy was indicated in another case in this group. Fortunately, laparoscopic suturing was achievably conducted to stop bleeding in both cases. For RALP group, there was 1 case discovered with postoperative complication from ureteral stent obstruction by blood clot. The patient was performed cystoscopy with stent changing under local anesthesia. Subsequently, there was no further postoperative complication found later.
In our research, we found that drain removal could be performed earlier in RALP group comparative to LP group with statistical signi cance. Median time for drain removal in RALP and LP groups were 3 (2-4) days and 2 (1-4) days, respectively. Since the RALP surgery encouraged a surgeon to certainly qualify anastomosis suturing procedure comparative to that of LP, the drain removal in RALP group could be conducted earlier. The results from meta-analysis performed by Light et al. [18] has found that LOH in RALP was shorter than that in LP for 1.2 days. In contrast, our research discovered that it was comparatively the same between both groups, 8 (5-15) days for LP versus 8 (6-13) days for RALP. The causes of longer LOH in our research was that the patients were regularly allowed to stay in hospital until drain and Foley catheter were off and patients get normal routine activity. Some patients were worried of the distance between the hospital and their residences and demanded to stay longer at the hospital to assure their recovery. In comparison with a study of Patel et al. [23], which has reported 1.1 days for LOH, whilst the recovery time until the patients could continue all routine independent activities of daily living was 7.7 days, this recovery time and our LOH were comparatively similar to each other.
Limitations found in our research were de ned. Since the study was performed with limited number of patients and the design of this study was retrospective review, then randomized control trial study was necessarily performed with more patients to con rm the results.

Conclusions
In era of minimally invasive surgery, LP and RALP treatments of UPJO have played critical roles and emerged widespread with equivalently quali ed success rates comparatively the same to that of the standard open technique. The robotic surgical systems have been continuously developed to overcome the limitations of laparoscopic approach. In our study, we have clearly indicated that RALP displayed higher advantages than LP in the features of shorter operative time, less blood loss and earlier drain removal. These merits assure us of using RALP as excellent, effective and harmless alternative treatment. Figure 1 Trocar placement for left robotic-assisted pyeloplasty. A = Assistant port (12 mm); C = Camera port (12 mm); R = Robotic arm port (8 mm).