The urological complications of renal transplantation: a series of 1535 patients

Authors


E. Streeter, Department of Urology, Churchill Hospital, Oxford, OX3 7LJ, UK.
e-mail: edwardstreeter@hotmail.com

Abstract

Objective  To determine the incidence of urological complications of renal transplantation at one institution, and relate this to donor and recipient factors.

Patients and methods  A consecutive series of 1535 renal transplants were audited, and a database of donor and recipient characteristics created for risk-factor analysis. An unstented Leadbetter-Politano anastomosis was the preferred method of ureteric reimplantation.

Results  There were 45 urinary leaks, 54 primary ureteric obstructions, nine cases of ureteric calculi, three bladder stones and 19 cases of bladder outlet obstruction at some time after transplantation. The overall incidence of urological complications was 9.2%, with that for urinary leak or primary ureteric obstruction being 6.5%. One graft was lost because of complications, and there were three deaths associated directly or indirectly with urological complications. There was no association with recipient age, cadaveric vs living-donor transplants, or cold ischaemic times before organ reimplantation, although the donor age was slightly higher in cases of urinary leak. There was no association with kidneys imported via the UK national organ-sharing scheme vs the use of local kidneys. The management of these complications is discussed.

Conclusion  The incidence of urological complications in this series has remained essentially unchanged for 20 years. The causes of these complications and techniques for their prevention are discussed.

Introduction

Major advances have been made over the last two decades in renal transplantation. While formal research programmes concentrate on the associated immunology technical issues concerning the procedure have almost been obscured. Despite this, early graft failure is often for technical reasons, and urological complications remain a major source of morbidity and occasional mortality. The incidence of these complications is discussed, and causes in their development evaluated. Recent advances in surgical technique are noted with speculation about their possible effect for the future.

Patients and methods

All cases of renal transplantation from the inception of our unit in January 1975 to May 1998 were included in the study; the series comprised 1535 consecutive renal transplants, in 1292 patients (mean age 43.0 years at surgery, range 11–75; male : female ratio of procedures 1.52 : 1, with 1386 cadaveric transplants and 149 living-related or unrelated donor transplants). All patients were followed up at the centre for at least 1 year after surgery. Data on the incidence and nature of urological complications were accumulated by retrospective case-note analysis.

Several operating surgeons were involved, of consultant or specialist registrar grade. In all but three procedures a Leadbetter-Politano ureteroneocystostomy was used, involving the passage of the ureter through a submucosal tunnel fashioned via a separate anterior cystostomy. Two cases required primary uretero-ureterostomy for a short donor ureter. In one procedure the ureter was implanted into a bladder caecoplasty. Ureteric stents were used rarely. A Foley catheter was used to drain the bladders of all patients for at least 5 days after surgery. Daily serum biochemistry was combined with careful clinical observation to monitor graft function. In recent years it has become standard practice to image all grafts by ultrasonography soon after surgery, usually in the first or second day, to detect early signs of vascular or urological complications. This succeeds the former practice of imaging only in those with suspected graft dysfunction. All patients since 1982 have been treated with cyclosporin-based immunosuppressive regimens. The 273 cases before this were treated with a combination of azathioprine and prednisolone, with high-dose prednisolone giving way to the current low-dose schedule in 1978.

For purposes of chronological comparison, data are subdivided into 1975–86 (transplants 1–600), 1986–91 (transplants 601–1000), and 1991–98 (transplants 1001–1535). Data relating to the early part of the series were published previously [1,2].

Results

Urinary leak

There were 45 cases of urinary leak, with a further three recurrent cases after a first operative repair. The median (range) time to onset was 29 (0–275) days, with all but two cases before 120 days. No grafts were lost, although one death could be directly attributed to subsequent sepsis. A summary of the position and timing of urinary leaks, with the subsequent management, is shown in Table 1.

Table 1.  Urinary leakage and obstruction; incidence and management
Location (n)Median (range) onset, daysTreatment (n)
Leakage
No site identified (1)14Death from cardiac arrest (1)
Upper ureter/collecting system (5)24 (0–49)Conservative (1)
  Retrograde stent (1)
Open closure of defect in renal pelvis (1)
Open closure of renal pelvis and stent (1)
Uretero-ureterostomy (1)
Lower ureter (14)21 (0–266)Nephrostomy (1)
Open drainage (1)
Open stent insertion (3)
Reimplantation (3)
Native ureteropyelostomy (2)
Uretero-ureterostomy (3)
Boari flap (1)
Bladder (6)47 (0–357)Prolonged catheter drainage (3)
Percutaneous drainage (1)
Repair of vesical defect (2)
Ureteric necrosis (19)35 (7–70)Reimplantation (5)
Native ureteropyelostomy (13)
Boari flap (1)
Obstruction
Upper/PUJ (10)350 (35–1610)Antegrade stent (4)
Ureteric dilatation (1)
Pyeloplasty (1)
Division of obstructing vessel (1)
Native ureteropyelostomy (3)
Middle (10)56 (1–1120)Antegrade stent (3)
Retrograde stent (1)
Open exploration (1) and stent (1)
Native ureteropyelostomy (2)
Uretero-ureterostomy (2)
Lower/VUJ (33)110 (1–2800)Nephrostomy (2)
Antegrade stent (6)
Balloon dilatation (2)
Endoscopic excision of suture (1)
Native ureteropyelostomy (2)
Reimplantation (20)

Minor leaks occurred early in four cases and were treated with re-catheterization or with observation alone. Three cases were managed initially with nephrostomy insertion, although two of these proceeded to ureteric reimplantation.

Thirty-seven cases required surgery; two with direct perforation of the ureter at procurement or implantation were sutured over. Three cases of vesico-ureteric leakage in the absence of necrosis were treated with open stent insertion. The degree of ureteric vascular compromise and necrosis dictated the use of reimplantation (eight cases), uretero-ureterostomy (three), uretero-pyelostomy (15) or Boari flap (two). Two cases of leakage through the anterior cystotomy required re-suturing of the bladder. Two cases required repeat reconstruction, with one leak persisting, being treated with nephrostomy and antegrade stenting. An additional operative failure was treated with antegrade stenting. Table 2 shows the relative incidence of urinary leakage over the study period.

Table 2.  The incidence of major complications with time
Transplant numberUrinary leak, %Ureteric obstruction, %
TotalIschaemic strictureOther causes
1–6002.804.331.832.50
601–10002.002.502.250.25
1001–15353.733.362.241.12
Total2.933.522.081.43

Primary ureteric obstruction

Ureteric obstruction with no external compression occurred in 46 patients; there were two groups, i.e. those with an ischaemic origin, becoming clinically evident usually at 1–18 months (32; median 6 months, range 0.5–47), and those where anatomical or technical factors were likely to have been more significant, evident in the early recovery period (14; median 3 days, range 0–11). There was no graft loss but there were two deaths, one caused by nephrostomy-related haemorrhage and the second with persistent stricturing, ureteric fistulation and sepsis. There were seven further cases of obstruction by lymphocoele (six) or obstructing blood vessels (one) in the early part of the series. Table 2 shows the relative incidence of ureteric obstruction over the study period.

The site and timing of obstruction again influenced management (Table 1). Antegrade stenting was used primarily in 19 cases, although five of these required subsequent operative procedures. Two distal strictures were successfully percutaneously dilated, but an attempt to remedy a pelvi-ureteric obstruction in the same way was unsuccessful, requiring surgery. One case of a very limited stricture at the ureteric orifice was successfully managed endoscopically via local excision.

An open operation was required in 33 cases. The ureter was unkinked in two cases and an overlying blood vessel divided in one. Proximal obstruction was treated by native ureteropyelostomy (five) or pyeloplasty (one). Distal obstruction favoured ureteric reimplantation in the 20 cases with a short segment stenosis, although one was later revised to a uretero-ureterostomy. Two further uretero-ureterostomies and one uretero-pyelostomy were performed for longer or mid-ureteric stenoses.

Transplant renal calculi

There were six cases of obstructive ureteric calculi, with three further cases not associated with obstruction. The median (range) time to presentation was 150 (56–1280) days. In the three unobstructive cases there was no intervention. Of those with obstruction, one was treated with nephrostomy insertion followed by percutaneous shock wave lithotripsy, three underwent successful percutaneous nephrolithotomy and one proceeded to open nephrolithotomy having failed the percutaneous approach. In the final case both endoscopic and percutaneous attempts failed to remove the calculus, with graft function being lost, resulting in a nephrectomy.

BOO

Seventeen patients developed significant BOO during the course of the study, about half presenting within the first month of transplantation. The causes were bladder neck stenosis (five), urethral strictures (four), BPH (five) and undetermined (three). Patients were treated with bladder neck incision (five), optical urethrotomy (two) urethral dilatation (two), TURP (five) or observation (three). There was one death after TURP caused by suprapubic catheterization, urinary extravasation and sepsis.

Haematuria causing obstruction

Six cases were recorded; of these, four occurred within 10 days of transplantation, causing hydronephrosis in two catheterized patients (days 2 and 3) and clot retention in two patients (days 6 and 8). One patient with hydronephrosis required cystoscopic irrigation, whilst the others were treated conservatively.

There were two cases each of ureteric obstruction caused by haemorrhage after percutaneous needle biopsy at 12 and 27 days, the latter associated with clot retention. Both were treated with nephrostomy.

Neoplasia

Two patients developed TCC of the bladder, both several years after transplantation. One had invasive disease, rapidly metastasising and leading to death. The other had superficial disease managed endoscopically. There was one case of nephrogenic adenoma at 4.5 years, excised endoscopically, recurring 2 years later and requiring two further cystoscopic resections. One patient who had had bilateral nephrectomies 3 and 4 years before transplantation for metachronous adenocarcinomas had widespread metastatic recurrence at 6 months and died.

Miscellaneous complications

Bladder calculi were found on three occasions, at 8 months to 4 years after surgery; in all three cases the calculi were removed cystoscopically. One further patient underwent cystoscopy for an ultrasonographically diagnosed bladder mass, which was found to be a protruding suture at the ureteric orifice, and which was excised endoscopically.

Association of complications with donor factors

The characteristics of the renal donors from the last 8 years were analysed to ascertain whether causal factors could be identified in the development of urological complications. In particular, the origin of the kidney, i.e. locally retrieved vs imported via the UK organ-matching scheme, the donor's age and the cold ischaemic time were recorded ( Table 3). There were no significant differences for the origin of the kidney, or cadaveric vs living donors. The donors were slightly older in cases with urinary leaks than in the overall population. There was a paradoxically shorter cold ischaemic time for ischaemic strictures.

Table 3.  Complications compared with donor factors
VariableTotal (535)Leak (20), %Ischaemic stricture (12), %
  • *

    P < 0.05;

  • P < 0.001.

Method of retrieval
Imported 149   4.7   3.4
Local 386   3.4   1.8
Donor
Cadaveric  55   5.4   1.8
Living-related 480   3.5   2.3
Mean (sd):
 donor age, years  41.25 (14.7)  45.8 (11.0)*  40.3 (14.5)
 cold ischaemic time, min1517 (488)1650 (598)1064 (286)

Discussion

Urological complications remain a major source of morbidity and occasional mortality in renal transplantation, despite a reduction in their incidence of at least half over the last 30 years. Table 5 [3–17] shows a comparison of the present with contemporary series including> 400 transplants. Similarly, the graft loss and related mortality has decreased, from 22% associated mortality and 54% graft loss in 1981 [3] to 3.3% and < 1%, respectively, reported here. The cause of these complications is of course multifactorial, and comparison of internationally published series shows wide variation among centres with different practices. We consider possible causal factors and how changes in practice may reduce the rate of complications further.

An initial dramatic reduction in urological complications at our unit (16% in the first 207 cases vs 5% in the next 400) was attributed to the change from high- to low-dose steroid-based immunosuppressive regimens, and the rate has changed little since then. As will be discussed, this may in fact mask an underlying trend towards safer surgery.

Causal factors in urological complications

Most urological complications are a result of technical errors at retrieval or reimplantation, or failure of tissue healing, influenced by ischaemia, inflammation, infection, immunosuppression and antiproliferative agents, and the nutritional state of the recipient. Four main categories of factors are considered, i.e. donor, recipient, technical and medical.

Donor

The donor's age and pre-existing comorbidity will influence the potential function of the transplanted organs, and their ability to withstand the insult of ischaemia and reperfusion. Once a decision is made to offer organs for transplantation, every effort must be made to preserve the donor in an optimal state, with periods of hypotension, inotrope support or prolonged stay in the intensive treatment unit all affecting the quality of the organs in the short- and possibly long-term. Retrieval surgery should thus proceed at the earliest appropriate opportunity.

The increasing pressure on transplant waiting lists will undoubtedly lead to more marginal donors being considered. The eventual effect of this tendency has yet to be evaluated, although in the last 8 years, the donor age for the present cases of urinary leak is slightly higher than for the whole population over the same period. Long-term graft survival is strongly correlated with donor-recipient HLA mismatch, e.g. in the data from the UNOS and UKTSSA registries [18,19]. This is of course a result of a combination of acute and chronic rejection. Loughlin et al.[5] reported no correlation of urological complications with HLA mismatch, but this series had an overall rate higher (13.2%) than that of most recently published series.

Recipient

The increasing availability of dialysis services and improved medical management of chronic renal failure have lead to an increase in the age of recipients of renal transplants in the present series, from a mean of 37.8 years in 1975–1986 to 46.7 years at operation in 1986–1998, although the present data showed no association between recipient age and complication rate. There was no association with primary disease and particularly no increased incidence of complications in diabetic patients.

Surgical

Technical considerations are of the utmost importance in the incidence of urological complications. Disruption of the ureteric normal blood supply at retrieval dictates that the remaining arterial and venous supply from the renal vessels and lower polar branches must be preserved by minimal peri-ureteric dissection, especially in the so-called ‘golden triangle’ between the ureter, kidney and renal artery. Even so, the distal ureter is prone to ischaemia. In living-related kidney donation, less aggressive preservation of the blood supply may be possible. In our unit 10% of transplants in this series were from living donors, with 12.5% of potentially ischaemic complications involved this subgroup. Thus no link is confirmed. This is in agreement with most large series of living-donor transplants, which report little effect on the incidence of urological complications [5,12].

Iatrogenic injury, the cause of two leaks in the present series, must be avoided during bench dissection. Minimizing warm ischaemic time is crucial. Dissection is carried out on ice and the organ may be contained in an ice-filled receptacle, e.g. a rubber glove, during the vascular anastomosis. This has been shown to maintain the core temperature of the kidney at < 10°C for protracted periods [20]. There was no association between cold ischaemic time or origin of the kidney (local vs from the national organ-sharing scheme) and the incidence of urological complication.

Ureteric reimplantation is where most interest has been focused in recent years, particularly for the type of vesico-ureteric anastomosis constructed and the use of prophylactic ureteric stents. In nearly all cases in the present series a Leadbetter-Politano submucosal tunnelled anastomosis was used [21]. This has been suggested to lower the incidence of ureteric reflux over conventional extravesical procedures [21,22], particularly important if the patient has recurrent UTIs. In the present series one case of symptomatic reflux was identified in a patient with recurrent pyelonephritis, successfully treated with ureteric reimplantation. More recent modifications of the extravesical approach, including a short muscular tunnel over the ureteric tip in an attempt to prevent reflux, have lead to the technique being shown in several series to produce fewer ureteric obstructions [6,23–25]. Ischaemic strictures are probably reduced through the shorter length of ureter required, and extrinsic compression from the submucosal tunnel is also avoided. The operating time is reduced and the need for an additional anterior cystotomy, the source of six leaks in our series, is obviated. The effect of this technique has been shown by two retrospective series comparing major complications before and after its adoption. butterworth et al.[25] claimed a reduction from 12% to 2% in a series of 248 patients, and Thrasher et al. [24] from 9.4 to 3.7% over 320 cases.

Two retrospective series showed remarkable reductions in major urological complications by using prophylactic ureteric stents (15% to 2.6%, and 13.6% to none, respectively) [26,27]. Several prospective randomized series have since confirmed this [16,28–30]. The stents are prone to breakage, especially if left for> 3 months, and may also migrate. The material used in the manufacture of the stent has been suggested to affect its lithogenicity [31], although the incidence of stent-associated calcification may be reduced by their early removal at 2 weeks after surgery [16]. The current policy in our unit (adopted after the conclusion of this series) is to use stents in cases where there is concern about ureteric ischaemia, combined with an extravesical anastomosis; it is too early to assess the effects of this recent change of practice.

In the last 535 patients the incidence of significant haematuria after transplantation, causing bladder outlet or ureteric obstruction, is low, with four cases within the first 10 days after surgery. Whether the source of the bleeding is from the cystotomy or elsewhere is unknown. All were managed conservatively with re-catheterization, one re-quiring cystoscopic irrigation. Two cases of obstructing haematuria secondary to allograft biopsy required nephrostomy insertion. All patients in our centre with cadaveric grafts undergo biopsies at 7 and 28 days, according to protocol or at any sign of deterioration of renal function. Living-related transplant recipients do not undergo routine biopsy because of the lower incidence of acute rejection in this group. These two cases thus represent a very low overall rate of urological complication, but of course do not include other cases of significant haemorrhage.

The incidence of lower urinary tract obstruction requiring a procedure within 6 months of transplantation was 2%. Whether to evaluate the lower urinary tract of asymptomatic patients before surgery, with the aim of reducing symptomatic outlet problems afterward, is an area of contention [32]. Again, whether to intervene beforehand, with the risk of stricturing of the urethra after instrumentation, without the regular passage of urine, is controversial [32]. Intermittent self-catheterization and instillation of antibiotic solution with normal voiding is suggested to circumvent this problem. Many investigators suggest waiting until after transplantation, with an early planned urological procedure. In the case of a noncompliant bladder requiring reconstruction or augmentation, enterocystoplasty is recommended before transplantation and the risk of immunosuppression [32].

Medical management

Of major importance was the change to low-dose steroid immunosuppression in the late 1970s, with the major complication rate decreasing from 16% in the first 207 cases to ≈ 7% since. Ongoing trials of immunosuppressive regimens are unlikely to produce this magnitude of effect. As has been stated, no clear link has been shown between rejection and urological complications, and as long as the complication is corrected, long-term function is preserved

Erectile dysfunction after transplantation is markedly less than in those with renal failure or dialysis, being halved from 47% to 22% between dialysis-dependent and successfully transplanted patients [33]. Ongoing problems are experienced often by patients on antihypertensive medication. The risk of erectile dysfunction after bilateral internal iliac transplant anastomoses has been estimated at 65%, vs 10% for unilateral cases [34].

Urological malignancy

Urological malignancy, like all other forms of malignancy in the transplant population, is in part a direct result of immunosuppression. An incidence of 1.4% (relative risk 7–11 times) was reported by Schmidt et al.[35]. However, all but 13 of the 868 patients reported in that series received antilymphocyte or antithymocyte globulin at induction. This practice, thought to be of great importance to the risk of developing malignancy, is very unusual in the present series. However, more recent data from the Australia and New Zealand Dialysis and Transplant registry confirmed a seven-fold increased risk of bladder and renal cancer in transplant recipients [36].

There were two cases of TCC of the bladder in the present series; invasive bladder cancer must be treated aggressively because of its propensity for rapid progression. For patients with superficial disease, management involving transurethral tumour resection and close cystoscopic observation should be tailored according to standard risk factors such as the presence of high-grade disease, multifocal invasion of the lamina propria or carcinoma in situ. Conventional adjuvant therapy using bCG is contraindicated in the immunosuppressed patient, thus early aggressive surgical treatment should be considered.

A particular risk factor for RCC is acquired renal cystic disease, which occurs in up to half of patients on long-term dialysis, often causing pain through cystic haemorrhage, haematuria and infection. It may be associated with a 30-fold increased risk of RCC (especially of the papillary variant) in the pretransplant population. The cystic change often regresses after successful transplantation but the relative risk of RCC in the transplant population is not thought to be high. There were no cases of RCC in the present series. Two prospective series of ultrasonography screening before transplantation reported incidences of occult tumours of 6% and 9% in patients with acquired renal cystic disease (1.5–2% in all patients presenting for transplantation) [37,38], but no survival benefit has been shown by screening. Our current practice is not to screen for the condition but to investigate urological symptoms according to standard practice.

The incidence of prostate cancer after renal transplantation has probably been underestimated because there is little screening. This may be of increasing importance with an ageing transplant population and the increasing longevity of grafts. The natural history of the disease in this population is unknown, especially important as the low serum testosterone levels often seen with chronic renal failure may be returned to normal with good renal graft function. The relative benefits of various treatment methods are likely to remain uncertain amongst transplant patients, and thus with no evidence to the contrary, they should be treated according to standard guidelines. Men age> 50 years, who may have no urinary symptoms if oliguric, should be considered for a DRE and serum PSA test before transplantation.

Conclusion

The major urological complication rate was 6.5% over the first 1535 cases of renal transplantation at our centre, and has remained constant for about the last 20 years. Causal factors were identified and possible changes in practice suggested, the results of which will be apparent over the next decade. With increasing experience of minimally invasive endoscopic and uroradiological techniques, the management of complications may be set to change simultaneously.

Authors

E.H. Streeter, MRCS.

D.M. Little, FRCS.

D.W. Cranston, FRCS.

P.J. Morris, PRCS.

Ancillary