Delayed proximal ureteric stricture formation after complex partial nephrectomy

Authors


Daniel Canter, Department of Surgical Oncology, Fox Chase Cancer Center, 333 Cottman Avenue, Philadelphia, PA 19111, USA. e-mail: danieljcanter@gmail.com

Abstract

Study Type – Therapy (case series)

Level of Evidence 4

What's known on the subject? and What does the study add?

As the indications for nephron sparing surgery continue to evolve, so do the potential complications.

This study examines a rare but likely underreported complication of nephron sparing surgery in order to better counsel and treat patients with complex renal tumours.

OBJECTIVE

  • • To report and review our incidence of delayed ureteric stricture (US) after complex nephron-sparing surgery (NSS).

PATIENTS AND METHODS

  • • Using our institutional kidney cancer database, we identified 720 patients who underwent NSS from 1 January 2000 until 31 December 2010 and identified eleven (1.5%) patients with a delayed US.
  • • Patient and tumour characteristics were reviewed.

RESULTS

  • • Median (range) tumour size and RENAL nephrometry score was 4.1 (2–7.2) cm and 10p (4–11p), respectively.
  • • There were eight of 10 solitary tumours (80%) located in the lower or mid-pole of the kidney.
  • • There were eight of 11 patients with delayed US (72.7%) who experienced a postoperative urinary leak.
  • • There were two of 11 (18.2%) patients who experienced a postoperative retroperitoneal haemorrhage, with one of these patients requiring selective embolization.
  • • All US were in the upper third of the ureter and were diagnosed at a minimum of 10 weeks postoperatively (median 154 days, range 70–400 days).

CONCLUSIONS

  • • US formation is an uncommon and under-reported event after complex NSS.
  • • Risk factors appear to include tumour complexity, imperative indications, mid- or lower pole location, postoperative urinary leak and haemorrhage.
  • • Although uncommon, postoperative US can occur after NSS for complex renal masses, necessitating patient counselling and diligent postoperative surveillance.
Abbreviations
NSS

nephron-sparing surgery

PN

partial nephrectomy

US

ureteric stricture.

INTRODUCTION

In the recent AUA guidelines concerning the clinical T1 renal mass, nephron-sparing surgery (NSS) is now the recommended standard of care for T1a enhancing renal masses and should be performed when anatomically feasible for T1b renal tumours [1]. Long-term oncological outcomes have shown that NSS has resulted in cancer-specific survival equivalent to that of radical nephrectomy for T1 renal masses [1–5], and emerging data suggest that NSS for pT2 renal tumours is not only technically feasible, but also oncologically sound [6,7]. The ability to maximize renal parenchymal preservation is essential for preventing chronic kidney disease [8,9] and its attendant deleterious overall and cardiovascular health effects [10–12].

Despite these advantages, both open and laparoscopic NSS are associated with the highest rates of major urological complications compared to other forms of therapy for localized RCC [1]. Large institutional series have shown that the rates of postoperative haemorrhage and urinary fistulae can range from 3.4–5.8% and 2.3–4.4%, respectively [13–15]. Furthermore, according to the AUA guidelines, major urological complications occur in 4.5–8.7% and 7.7–10.6% of patients undergoing open and minimally-invasive partial nephrectomy (PN) [1]. Late complications of surgery are often the least well documented and therefore often under-reported. As such, we are aware of only one previous single institutional report evaluating the development of a ureteric stricture (US) formation after partial nephrectomy [16].

In the present study, we report our experience with the delayed complication of US after complex NSS. Pre-, peri- and postoperative data are reviewed to delineate the potential risk factors for the development of a delayed US after complex PN.

PATIENTS AND METHODS

We queried our prospectively maintained institutional kidney tumour database and identified 720 patients who were surgically treated by partial nephrectomy for solid renal masses from 1 January 2000 until 31 December 2010. Of the 720 patients who underwent PN, 416 (57.8%) patients were treated with an open PN, 234 (32.5%) patients underwent a robotically-assisted PN and 70 (9.7%) patients were treated with a laparoscopic PN. In total, eleven (1.5%) patients with delayed proximal ipsilateral US were identified. Their records were reviewed to identify pre-, peri- and postoperative characteristics that may have predicted or contributed to the development of a delayed ureteric complication.

Demographic factors such as age, sex, indication for NSS, estimated blood loss, ischaemia time, postoperative complications, tumour size, tumour location and nephrometry score (largest nephrometry score for patients with multifocal tumours) were evaluated and analyzed. The RENAL nephrometry scoring system is based on the five most reproducible features that characterize the anatomy of a solid renal mass: (R)adius (scores tumour size as maximal diameter), (E)xophytic/endophytic properties of the tumour, (N)earness of the deepest portion of the tumour to the collecting system or renal sinus, (A)nterior (a)/posterior (p) descriptor and the (L)ocation relative to the polar line. All components except for the (A) descriptor are scored on a 1-, 2- or 3-point scale. The (A) describes the principal mass location to the coronal plane of the kidney. The suffix ‘x’ is assigned to the tumour if an anterior or posterior designation is not possible. An additional suffix ‘h’ is used to designate a hilar location of the tumour (abutting the main renal artery or vein). Masses with a RENAL nephrometry score totalling 4–6 were considered low complexity for resection, those with a score of 7–9 were considered moderate complexity, and those with a score of 10–12 were considered high complexity [17].

RESULTS

Overall median (range) age, maximal tumour diameter and RENAL nephrometry score was 61 (51–79) years, 4.1 (1.4–7.2) cm and 10p (4–11p), respectively. Overall median follow-up from initial surgery was 23 months. There were eight male patients and seven tumours located on the right kidney (Table 1). There were five patients with a tumour located posteriorly, two patients with a tumour located anteriorly, one patient with a tumour located anteriorly and hilar, and three tumours were denoted as ‘x’, signifying that they were symmetrically placed across the axial midline. There were four tumours each located in the lower pole or mid-pole, whereas two tumours had an upper pole location. A single patient had multifocal tumours. Of the 11 patients, eight had an imperative indication for NSS: three patients with chronic kidney disease stage III or higher, one patient with a solitary kidney, two patients with a solitary kidney and chronic kidney disease stage III, and two patients with bilateral renal masses (including one with multifocal disease) and chronic kidney disease stage III. A single patient had a low complexity tumour, four tumours were considered moderate complexity as defined by nephrometry score, and six patients had a high complexity tumour.

Table 1.  Patient clinical and demographic characteristics
CharacteristicValue
Age (years) 
 Mean (sd, range)63 (7.9, 51–79)
 Median61
Sex, n (%) 
 Men8 (72.7)
 Women3 (27.3)
Tumour side, n (%) 
 Right7 (63.6)
 Left4 (36.4)
Maximal tumour diameter (cm) 
 Mean (sd, range)4.4, 2 (2–7.2)
 Median4.1
Nephrometry score 
 Median (range)10p (4–11p)
Tumour location, n (%) 
 Coronal plane 
  Anterior (a)2 (18.2)
  Anterior (a) and hilar (h)1 (9.1)
  Posterior (p)5 (45.5)
  No a/p designation (x)3 (27.3)
 Polar line 
  Upper2 (18.2)
  Mid4 (36.4)
  Lower4 (36.4)
  Multifocal1 (9.1)
Preoperative serum creatinine (mg/dL) 
 Mean (sd, range)1.61 (1.2, 0.8–4.6)
 Median1.4
Preoperative estimated GFR (mL/min) 
 Mean (sd, range)53.5 (20.3, 12–80.4)
 Median54.8
Preoperative chronic kidney disease stage, n (%) 
 I1 (9.1)
 II3 (27.3)
 III6 (54.5)
 IV–V1 (9.1)

Operative details are presented in Table 2. In total, 10 patients underwent an open partial nephrectomy and one patient had a robotically-assisted laparoscopic partial nephrectomy. There were two patients who also underwent a segmental renal vein thrombectomy at the time of the PN. In all, two patients had JJ ureteric stent placed at the time of the initial tumour excision. Renal artery ischaemia was used in 10 patients; cold ischaemia was employed in 8 of these cases to decrease ischaemic injury. Median (range) ischaemia time was 43.6 (15–64) min for the entire group and 46 (19–60) min for the cold ischaemia group. There were two patients who had multiple renal arteries and/or renal veins. A single patient had a crossed-fused ectopic kidney. Median (range) estimated blood loss was 150 (50–1500) mL. In all, nine patients had RCC confirmed on pathology, with clear cell RCC being most common (7 tumours). All surgical margins were negative.

Table 2.  Surgical and pathological characteristics
CharacteristicValue
Type of surgery, n (%) 
 Open10 (91.9)
 Laparoscopic0
 Robotic1 (9.1)
Vessels clamped, n (%) 
 None1 (9.1)
 Renal artery only1 (9.1)
 Renal artery and renal vein7 (63.6)
 Renal artery (multiple) and Renal vein (multiple)2 (18.2)
Type of ischaemia, n (%) 
 None1 (9.1)
 Cold8 (72.7)
 Warm2 (18.2)
Ischaemia time (min) 
 Mean (sd, range)40.3 (18.6, 15–64)
 Median43.6
Estimated blood loss (mL) 
 Mean (sd, range)286.4 (413, 50–1500)
 Median150
Pathological distribution, n (%) 
 Chromophobe RCC1 (9.1)
 Clear cell RCC7 (63.6)
 Oncocytoma2 (18.2)
 Papillary RCC1 (9.1)
Pathological stage, n (%) 
 T1a4 (36.4)
 T1b1 (9.1)
 T2a2 (18.2)
 T2b0
 T3a2 (18.2)
 T3b2 (18.2)

In total, eight patients had a postoperative urinary fistula (Table 3). There were four patients who had an immediate postoperative fistula, diagnosed by increased drainage and an elevated retroperitoneal drain creatinine. Of the four patients with a delayed urinary fistula, two patients presented with ipsilateral flank pain (on postoperative days 45 and 146, respectively), one patient presented with generalized abdominal pain (on postoperative day 32) and one patient presented incidentally on routine postoperative imaging for staging purposes before starting targeted therapy (on postoperative day 46). Of the eight patients with postoperative urinary fistula, six had high complexity tumours by nephrometry score, and the remaining two patients had moderate complexity tumours. There were four patients who were managed with a combination of a retroperitoneal drain and an indwelling ureteric stent; two patients were managed with indwelling ureteric stents only, one patient was managed with a percutaneous nephrostomy, and one patient was managed with a retroperitoneal drain. There were two patients who developed a large retroperitoneal bleed, with one patient requiring selective renal artery embolization.

Table 3.  Postoperative leak and ureteric stricture characteristics
CharacteristicValue
  1. ARF, acute renal failure; PCN, Percutaneous nephrostomy; RPD, retroperitoneal drain.

Postoperative leak 
 Patients (n)8
 Immediate leak, n (%)4 (50.0)
 Delayed leak, n (%)4 (50.0)
 Postoperative days to leak 
  Mean (sd, range)33.6 (49.9, 0–146)
  Median16
 Treatment 
  JJ, n (%)2 (25.0)
  PCN, n (%)1 (12.5)
  RPD, n (%)1 (12.5)
  JJ and RPD, n (%)4 (50.0)
Proximal ureteric stricture 
 Patients (n)11
 Presenting symptom/sign 
  Ipsilateral flank pain, n (%)2 (18.2)
  ARF, n (%)5 (45.5)
  Incidental hydro on imaging, n (%)1 (9.1)
  Incidental stricture on imaging, n (%)3 (27.3)
 Postoperative days to stricture 
  Mean (sd, range)172 (89.7, 70–400)
  Median154
 Initial treatment 
  JJ8 (72.7)
  PCN3 (27.3)
 Other treatments 
  Endoscopic dilation5 (45.5)
  Ileal ureter1 (9.1)
  Completion nephrectomy1 (9.1)

All US were proximal and were diagnosed a minimum of 10 weeks postoperatively (median 154 days, range 70–400 days) (Table 3). Only two patients presented symptomatically with flank pain. There were five patients who presented with acute renal failure and four patients were diagnosed incidentally on routine imaging. In all, eight patients were treated initially with an indwelling JJ ureteric stent, whereas three patients were managed with a percutaneous nephrostomy tube. There were five patients who underwent endoscopic treatment of their stricture, including ureteric dilation, laser ureterotomy or a combination of the two procedures. A single patient had a completion nephrectomy secondary to a continued urinary fistula, recurrent infections and only ≈10% function of the affected kidney. A single patient underwent definitive repair with an ileal interposition. The remaining seven patients have been managed with a chronic indwelling JJ ureteric stent.

DISCUSSION

Multiple single institutional series have established NSS as an oncologically equivalent approach to localized RCC at the same time as preserving renal parenchyma [1–3,18,19]. Recent AUA guidelines state that recurrence-free survival for T1 renal masses treated by PN or radical nephrectomy are both greater than 98.0% [1]. At the same time, the AUA guidelines also show that both open and minimally-invasive PN have the highest rates of major urological complications, specifically retroperitoneal haemorrhage and urinary fistula. US are a rare and most certainly under-reported complication of NSS given their delayed presentation. To date, only one previous report in the literature exists. In the present study, we identify an incidence of delayed US of 1.6%. All strictures were proximal and all occurred in a delayed fashion. A review of the records of the 11 affected patients shows that tumour complexity, mid-/lower pole tumours, imperative indications, urinary leak and haemorrhage were all associated with the development of a delayed US.

Only one other series has reported the development of delayed US after PN. Singh et al. [16] described their experience with an incidence of 0.35% (four of 1154 cases). In their series, all cases were via an open approach, and the tumour size range was smaller than that in our present series (2.1–4.5 cm), reflecting the ever expanding indications for NSS. Similar to our experience in the present study, three patients had a lower pole tumour, and one patient had an anterior medial interpolar tumour. Furthermore, one patient had a retroperitoneal bleed that required reoperation. Time to diagnosis of US in their series was 7 weeks. Singh et al. [16] suggest that the development of a delayed US was attributable to either ischaemic or mechanical injury of the proximal ureter during dissection of these inferiorly positioned tumours at the same time as denuding the perinephric fat. Furthermore, they postulate that ligation of the ipsilateral gonadal vein may contribute to ischaemia of the proximal ureter leading to stricture formation.

Our present series included 10 patients who underwent open excision of their renal tumour and one patient who underwent a robotically-assisted tumour excision. By contrast to findings of the study Singh et al. [16], tumour size in our present series was in the range 2.0–7.2 cm. Although ischaemia and/or mechanical injury to the proximal ureter are plausible explanations for delayed stricture formation, the fact that three-quarters of the proximal US in our cohort were preceded by a urinary fistula appears to indicate that the inflammatory process in the retroperitoneum caused by urinary extravasation may contribute to an inflammatory substrate leading to stricture formation. In addition, we aimed to determine whether there was a correlation between the presence of an accessory lower pole or multiple renal arteries and the development of a delayed US. In our series, two patients had multiple renal arteries and one patient had a cross-fused ectopic kidney with aberrant renal vasculature. In our database as a whole, 24.6% of PNs were performed on renal units with greater than one renal artery, which is comparable to the rate of multiple renal arteries in patients who developed a US.

NSS has evolved into the standard of care to treat T1a and T1b renal masses as a result of its advantage of renal preservation without sacrificing oncological principles. However, this benefit is set against the increased risk of major urological complications compared to radical nephrectomy, active surveillance or thermal ablation [1]. Fortunately, the development of proximal US after NSS is rare, with a total of 15 cases being reported in the world literature. Nevertheless, this is a cause of substantial morbidity in the affected patient. In our present cohort, this rare complication occurred in a group primarily comprising patients with imperative indications for PN; thus, the morbidity of a US is relatively small compared to the risks of death and hospitalization from chronic kidney disease and dialysis [20,21].

The present study is limited by its retrospective nature and the inherent tendency to generate a causation for an effect. Although, in our present series, there appeared to be a link between the development of US after complex NSS for a high complexity tumour located in the lower or mid-pole where a urinary fistula developed postoperatively, we also had a large number of lower/mid-pole PNs for high complexity tumours where a US did not occur. Moreover, in our dataset, urinary fistula in general is infrequently associated with US formation.

Partial nephrectomy is increasingly being used to resect more complex renal tumours in an attempt to preserve renal function. As the indications for PN continue to evolve, so do the potential complications. Ipsilateral US formation is an uncommon, often delayed (minimum of 10 weeks in the present series) and proximal in location. The potential for this complication must be monitored. There is a paucity of data regarding potential delayed surgical complications. In the case of complex NSS, the 1–2% risk of delayed US should be considered in the risk/benefit equation when determining the optimal treatment strategy for each individual patient. Careful postoperative follow-up, especially in more complex tumours, must be initiated to limit morbidity.

ACKNOWLEDGEMENTS

The authors would like to thank Debra Kister and Michelle Collins for their management of the Fox Chase Kidney Cancer Keystone Database.

CONFLICT OF INTEREST

This publication was supported in part by grant number P30 CA006927 from the National Cancer Institute. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health. Additional funds were provided by the Fox Chase Cancer via institutional support of the Kidney Cancer Keystone Program.

Ancillary