Laparoscopic Retroperitoneal Dismembered Pyeloplasty: Single-Center Experience in China

Introduction

In 1993, laparoscopic pyeloplasty was initially reported by Schuessler et al. ¹ Since then, it has been gradually recognized as a safe and less invasive reconstructive procedure for ureteropelvic junction (UPJ) obstruction compared with open pyeloplasty. The success rate of the laparoscopic operation has been reported to be the same as for open procedures. ² Moreover, the principal difference between laparoscopic pyeloplasty and other endourologic procedures lies in its ability to treat all patients with UPJ obstruction. ³ In the present article, we report on our experience with laparoscopic retroperitoneal dismembered pyeloplasty in the treatment of UPJ obstruction. Special attention was paid to the technical features associated with the procedure.

Subjects and Methods

From May 2004 to April 2012, in total, 117 consecutive patients (64 men and 53 women) with a mean age of 33.5 years (range, 12–60 years) underwent laparoscopic retroperitoneal dismembered pyeloplasty for UPJ obstruction at our institution. Obstruction was present on the right side in 57 patients and on the left in 60. Seventy-eight patients were symptomatic; in the 39 patients who were asymptomatic, UPJ obstruction was detected incidentally by renal ultrasonography or computerized tomography. Five patients had undergone previous unsuccessful endourological procedures. One patient had a congenital solitary kidney. Nine patients had renal calculi, including horseshoe kidney in 4 patients. UPJ obstruction was confirmed by diuresis renography, intravenous urography, and retrograde pyelography in all cases.

The patient was placed in the full flank position under general anesthesia. A four-port blunt finger or balloon-dissecting retroperitoneal laparoscopic approach was used as described previously. ⁴ The UPJ was left unstented in order to confirm the natural position of the obstructed ureter and crossing vessels during laparoscopy. Initially, the ureter was identified and then dissected upward to reach to the pelvis. The proximal ureter and the renal pelvis, including the UPJ, were fully mobilized, and the cause of UPJ obstruction (crossing vessels, stenosis) was evaluated. According to the intraoperative findings, the redundant renal pelvis was reduced, and the stenotic segment of the UPJ was excised. If anterior crossing vessels were present, the ureter was transposed anterior to the vessels. In cases of calculi in the collecting system, the stones were removed directly from the renal pelvis if visible, or a flexible cystoscope was passed through a laparoscopic trocar to allow for stone extraction via pyeloscopy.

The pelvis was longitudinally incised, and the ureter was spatulated to sufficient length while keeping the remaining continuity between pelvis and ureter intact at this time. After the deepest point of the renal pelvis was sutured to the apex of the spatulated ureter (the first suture was left untied in cases of crossing vessel), the ureter was dismembered from the pelvis, and both were trimmed further. A free-hand laparoscopic running anastomosis was performed with two separate 4-0 polyglactin sutures. The posterior ureteropelvic anastomosis was completed with the first suture up to the proximal end of the ureter. Then, a double pigtail stent was inserted directly in an antegrade fashion through the assistant trocar from the outside of the body with the assistance of a guide wire. The anterior ureteropelvic anastomosis was subsequently closed with the second suture, from the lowest point of the anterior wall up to the top of the pelvic opening. The second suture was then tied to the first suture at the proximal end of the ureter. A drain was placed through the trocar in the midaxillary line into the perinephric space adjacent to the anastomosis.

The urethral catheter was removed on postoperative Day 2 if there was minimal output from the retroperitoneal drain. The drain was removed when the drainage was less than 15 mL within 24 hours. The ureteral stent was removed 6 weeks postoperatively. Diuresis renography was performed at 3 and 6 months postoperatively and annually thereafter. Success was defined as symptomatic relief and improvement of the diuresis renogram without evidence of obstructed drainage. A retrograde pyelography was performed if the diuresis response was equivocal. Outcome data, including operative time, blood loss, conversion, complications, length of hospital stay, and follow-up renogram, were prospectively collected and recorded.

Results

There was no conversion to open surgery. All operations were completed laparoscopically. The mean operative time was 170 minutes (range, 90–310 minutes) for all cases: 210 minutes for the first 28 cases and 150 minutes for the last 89 cases. The mean estimated blood loss was 40 mL (range, 20–100 mL), and no patient required blood transfusion. Crossing vessels as the reason of the UPJ obstruction were noted in 18.0% of the patients (21/117). In all cases, the ureter was transposed anteriorly. Coexisting renal calculi were removed completely in all 9 cases, including horseshoe kidney in 4 cases. The mean hospital stay was 7 days (range, 5–14 days). Two patients had urine leak at the anastomosis on postoperative Day 1, which was documented via the retroperitoneal drain; after a prolonged duration with the drain on suction in the retroperitoneal space (14 days and 12 days, respectively), the urine leak ceased spontaneously. There were no other complications during the intraoperative or postoperative period.

The mean follow-up was 43 months (range, 3–95 months). Seven patients developed anastomotic strictures, which required open surgical repair or ureteral stent. In all other patients, the obstruction was resolved or significantly improved with the diuretic renogram and radiological study with a success rate of 94% (110/117).

Discussion

Laparoscopic pyeloplasty can be classified into two types by approach: transperitoneal and retroperitoneal. Most reports on laparoscopic pyeloplasty concerned the transperitoneal approach in which some authors emphasized that it offered more working space and a better field of view. ⁵ However, we prefer the retroperitoneal approach because the majority of our laparoscopic procedures at our institution are performed through the retroperitoneal route, and we were familiar with this surgical technique. The retroperitoneal approach has several advantages. First, it is a more direct approach, and the exposure of the pelvis is excellent. Less dissection is required to expose the UPJ, which can reduce the operative time and minimize risk of injury to intraperitoneal organs. Second, the potential harmful effects of peritoneal exposure to blood, urine, and carbon dioxide are avoided. Third, a retroperitoneal approach is feasible even in patients who have undergone prior abdominal surgery. Furthermore, if the retroperitoneal procedure is unsuccessful, it is still possible to perform a second surgery by the transperitoneal approach because intraperitoneal structures are not affected.

Although it has been stated that transposition of the UPJ anterior to lower pole vessels is easier using the transperitoneal route, in our experience, we have not found transposition of the UPJ difficult during the retroperitoneal approach. We believe that the retroperitoneal laparoscopic approach has greater sensitivity for detecting a crossing vessel because it allows the surgeon a view of the UPJ and related vessels in their anatomic position from laterally to medially by simple elevation of the lower pole of the kidney. In spite of these controversies, we support the view that the preferred approach for laparoscopic pyeloplasty will mainly depend on previous experience and training of the individual surgeon. ⁶

Although the diagnosis was confirmed preoperatively, it is very important to make a plan on the basis of intraoperative findings. We advocate that surgeons should avoid placing a retrograde ureteral stent at the beginning of the procedure in order to take advantage of dilation of the renal pelvis. Without a stent in place, the renal pelvis invariably remains distended, which helps in evaluating the cause and precise location of the obstruction. The distended pelvis also helps in the dissection of the pelvis, especially in cases of peripelvic adhesions. Furthermore, laparoscopic stenting in our experience saves on the additional operative time required for cystoscopic placement of the ureteral stent and changing the patient's position on the operating room table. In antegrade placement, the stent is inserted only after completion of the posterior anastomotic sutures in order to avoid interfering with posterior wall suturing, as it is done during the open procedure. Through the assistant trocar, we placed the double pigtail stent directly into the ureter from the outside of the body with the use of a guide wire and then verified the stent's position in the bladder by observation of the reflux of saline containing methylene blue from the bladder. We find this method simple, efficient, and straightforward.

There are various types of pyeloplasty, including Anderson–Hynes dismembered, Fenger plasty, and Y-V plasty, performed laparoscopically through a transperitoneal or retroperitoneal approach. ⁷ We routinely performed dismembered pyeloplasty in order to completely duplicate the fundamental principle of the traditional open surgery. Dismembered pyeloplasty remains the gold standard for surgical treatment of UPJ obstruction, with a long-term success rate exceeding 90%, as this technique involves the complete removal of both the anatomical and functional causes of obstruction (stenosis, adynamic junction). ⁸ There is now evidence in the laparoscopic literature that a nondismembered technique such as Fenger plasty leads to a lower success rate than a dismembered pyeloplasty. ³ A reasonable explanation for this finding is that a Fenger plasty does not eliminate the adynamic segment at the UPJ that causes the obstruction and does not resect the redundant renal pelvis. Some authors have reported that resection of a redundant renal pelvis is not always required for UPJ obstruction. ⁹ They consider a redundant renal pelvic wall useful in obtaining significant length to anastomose the renal pelvis and ureter without tension. It is, however, our experience that reduction of any redundant renal pelvis in conjunction with complete resection of the stenotic ureteral segment improves the urodynamics of the new UPJ and the transport of urine down to the ureter with a 94% success rate using the laparoscopic Anderson–Hynes technique. Our results are similar to those reported by Zhang et al. ¹⁰

Up to now, most of the surgeons still admit that the disadvantage of the laparoscopic pyeloplasty is the long operative time. The reconstruction portions of the procedure take the most time because it requires advanced laparoscopic techniques to completely dismember the UPJ, spatulate the ureter, and complete an intracorporeal laparoscopic anastomosis. In the current series, the mean operative time was 210 minutes for the first 28 cases and 150 minutes for the last 89 cases, since we began more complicated laparoscopic procedures, such as cystectomy and prostatectomy, at our institution. These results suggest that as with most advanced laparoscopic procedures, there is a learning curve associated with laparoscopic dismembered pyeloplasty. In our experience, there were several important technical points that helped to reduce operative time. First, antegrade placement of a ureteral stent as previously described reduced operative time by approximately 20–30 minutes. Second, keeping the continuity between the pelvis and ureter at all times is very important. For a novice surgeon, lateral spatulation of the ureter after dismembering the UPJ is difficult, and rotation of the ureter during spatulation is always a risk. Keeping the continuity of the UPJ provides the surgeon a sense of direction for accurate and linear spatulation. Lastly, use of a running anastomotic suture is our preferred technique for the pelviureteral anastomosis. This technique only requires two sutures and four to five knots tied. Our stepwise approach proved to be both reproducible and teachable at our institution, with all three technical modifications contributing to overall reduction in operative time.

It is evident that laparoscopic pyeloplasty was applicable to secondary UPJ obstruction and other complex cases associated with renal calculi or congenital anomalies such as horseshoe kidney and solitary kidney, with results equivalent to those of open surgery. ¹¹ In the current series, primary treatment elsewhere had failed in 5 patients, including balloon dilation in 2 and endopyelotomy in 3. The mean operative time was 20 minutes longer in the secondary pyeloplasty group compared with the primary cases; however, there was no increase in blood loss or complications. Through the retroperitoneal approach, renal calculi were easily extracted in 9 patients, and all patients were rendered free of stones. Of the 9 patients with stones, there were 4 cases of horseshoe kidney with a pelvic stone. In the literature, UPJ obstruction occurs in approximately 15%–33% of horseshoe kidneys. ¹² Application of laparoscopic pyeloplasty to UPJ obstruction in the horseshoe kidney has been limited to the transperitoneal approach to date because a wide working space was thought to be required for this unique anatomic situation. ¹³ In our 4 cases, a retroperitoneal approach was used with all four trocars placed more medially than normal. Following sufficient dissection of kidney and distended renal pelvis, we found the working space was more than adequate. The intraoperative and postoperative outcomes appear to be satisfactory. Our results show that the success rate in these complex cases was 100%; therefore we considered that the indication of laparoscopic retroperitoneal dismembered pyeloplasty includes even these more challenging situations.

Conclusions

In summary, there is no doubt that laparoscopy has become an important surgical tool in the treatment of many urological diseases. Laparoscopic retroperitoneal dismembered pyeloplasty for the repair of UPJ obstruction provides excellent results similar to those of open pyeloplasty with less morbidity. It is suitable for both primary and secondary UPJ obstruction, with crossing vessels, renal stones, or congenital anomalies such as horseshoe kidney. Our results demonstrate that laparoscopic retroperitoneal dismembered pyeloplasty is safe, effective, and technically feasible. However, the reconstructive step of the operation is technically challenging and required advanced laparoscopic skills. The length of the learning curve depends on the number of procedures performed and their frequency. With the improvement of the technique, we believe that it would become a new standard treatment for UPJ obstruction.