Objectives: The aim of our study was to evaluate the safety and efficacy of percutaneous treatment of ureteral obstructions and leak after renal transplant and to evaluate the long-term results and graft survival rates in a single center.
Materials and Methods: This retrospective study included 27 transplant recipients who received percutaneous treatment between January 2000 and December 2010 and who had follow-up data until December 2018. During this period, 294 renal transplants were performed at our institution, with 17 (5.7%) having a ureteral complication. Ten patients included in the study had their transplants at another center. Percutaneous nephrostomy, balloon dilatation, and double J stent placement were used in the management of complications. Cutting balloon dilatation and tandem ureteral stent placement were done in cases of resistant stenosis. Technical success and ureter patency rates were calculated. Graft survival rates were compared between early and late obstruction groups and between successful and unsuccessful interventional treatment.
Results: Among included cases, 21 obstructions (7 early, 13 late) and 8 leaks were detected. The technical success rate of percutaneous nephrostomy was 100% in all groups. The technical success rates of balloon dilatation and double J stent were 100% and 88% in the early and late obstruction groups, respectively. Censored graft survival rates in all groups at 1, 5, and 10 years were 89%, 89%, and 73.9%, respectively. In long-term follow-up, ureter patency rates were 100%, 33%, and 50% for early obstruction, late obstruction, and urinary leak groups, respectively (P = .018). Graft survival rates between early and late obstruction groups were not significantly different. No major complication, allograft loss, or 30-day mortality was seen.
Conclusions: Percutaneous management of ureteral complications is safe and effective and should be considered as first-line treatment because of its less invasive nature and lower complication and morbidity rates.
Key words : Interventional radiology, Kidney transplantation, Ureter complications
During the first years after renal transplant had started, urological complications were shown to occur at rates of 10% to 25%; however, with development of new neocystostomy techniques, the present rates of occurrence have decreased to 1% to 8%.1 The most common urologic complications are ureteral obstruction and leak. Although their incidence has declined, they still remain one of the main causes of morbidity and may cause graft loss in renal transplant recipients. Traditionally, surgical revision is used to treat these complications; however, interventional radiology also plays a major role in the diagnosis and treatment of postoperative complications.2 The aim of our study was to evaluate the safety and efficacy of percutaneous treatment of ureteral obstructions and leak after renal transplant and to evaluate the long-term results and graft survival rates in a single center.
Materials and Methods
Between January 2000 and December 2010, there were 27 renal transplant recipients (18 males, 9 females; age range, 8-64 y; mean age of 34 y) who had urologic complications and were treated percutaneously in our hospital’s interventional radiology unit; these patients had follow-up data to December 2018. After approval from our institutional review board, the data were obtained retrospectively from medical records. We collected patient age and sex, donor type, time interval between transplant and complication, type and level of complication, the initial percutaneous procedures, and, if present, recurrences and follow-up data. Patients were classified as those with early obstruction (≤ 3 mo posttransplant), late obstruction (> 3 mo posttransplant), and urinary leak. In addition, in the obstruction groups, preprocedural and postprocedural creatinine values were recorded. Eleven patients received grafts from living donors, and 16 received grafts from deceased donors. Seventeen transplants were conducted in our transplant unit with a ureteral anastomosis using the Lich Gregoir technique, and 10 were done in another center (Table 1).
Patients with urinary obstruction presented with decreased levels of urine output and increased levels of serum creatinine and had hydronephrosis and ureteral dilatation on ultrasonography.
Patients with urinary leak presented with decreased levels of urine output, anuria, fever, and urine drainage from the surgical drain if present. On ultrasonographic examination, no incidences of hydronephrosis were detected and levels of urinary leakage were determined with antegrade pyelography.
Informed consent was obtained from all adult and pediatric patients (through their guardian) before intervention procedures. Platelet count and prothrombin time (international normalized ratio) values were normalized if needed to reduce the risk of bleeding during and after the procedure. All patients received prophylactic broad-spectrum antibiotics before the procedure.
The first-step option for treatment was percutaneous nephrostomy (PCN) to allow renal function to recover and for any associated ureteral leaks to heal. All procedures were performed under ultrasonographic and fluoroscopy guidance with moderate intravenous sedation. In kidney grafts with hydronephrosis, an 18-gauge needle was used for percutaneous puncture. Through the needle, a guidewire (Amplatz Super Stiff, Boston Scientific, Natick, MA, USA) was placed. In kidney grafts with no or mild hydronephrosis, percutaneous puncture was performed with a 21-gauge needle under ultrasonographic guidance, and an antegrade pyelogram was performed to confirm the diagnosis. Over a 0.018-inch nitinol guidewire, a coaxial introducer (AccuStick Introducer System, Boston Scientific) was placed. Next, over a stiff guidewire (Amplatz Super Stiff, Boston Scientific), an 8F- to 10F-nephrostomy drainage catheter was placed into the renal pelvis.
Percutaneous balloon dilatation (BD), double J stent (DJS) placement, and other interventional procedures were performed within 1 to 30 days after PCN according to the patient’s clinical status. The stenotic segment or leakage site of the ureter was passed through the nephrostomy tract using a 0.035-inch hydrophilic wire (Terumo, Tokyo, Japan). In patients with ureteral obstruction, after the hydrophilic wire was exchanged with a stiffer guidewire, the stenotic segment was dilated with a 7- to 8-mm and 4-cm-long high-pressure balloon (Bluemax, Boston Scientific). After BD, a DJS of 8F and 14 to 16 cm (Mar Flow Renal Transplant Stents, Zurich, Switzerland) was placed with one end in the bladder and the other in the renal pelvis, and a control 8F nephrostomy catheter was left clamped in place in the renal pelvis for 24 hours. After 24 hours, a control pyelogram was performed to confirm ureter patency, and then the nephrostomy catheter was removed. Double J stents were removed within 1 to 6 months in accordance with the patient’s clinical status.
In patients with resistant stenosis, cutting balloon dilatation (CBD) with a peripheral balloon device (Boston Scientific) was performed and tandem ureteral stents (TUS) were placed. Tandem ureteral stents were placed separately over 2 guidewires. In patients with ureter leak, perirenal collections were drained percutaneously if needed.
All patients were followed with measurement of creatinine levels and ultrasonographic examinations. All complications were graded according to the quality improvement guidelines of the Society of Interventional Radiology.3
Technical success was defined as the ability to place PCN and crossing the level of obstruction and to place DJS with or without BD during the procedure. Ureter patency was defined as improvement in the obstruction site and achieving ureteral patency without the need of a DJS and resolution of the symptoms of obstruction. In the urinary leak group, ureter patency was defined as disappearance of the leak and achievement of ureteral patency without the need for a DJS. Graft loss was determined as the date a patient returned to hemodialysis/peritoneal dialysis or had a graft nephrectomy. Graft survival was measured from PCN placement until last follow-up or graft loss. Graft survival was measured only in patients in whom ureteric obstruction and leak were the cause of transplant dysfunction; therefore, patients who had biopsy-proven rejection on initial administration or after PCN were excluded from graft survival results. Patients who underwent surgical treatment immediately after PCN and patients who had unsuccessful interventional treatment in long-term follow-up were classified in the same group for graft survival.
Serum creatinine levels were compared with t tests for early and late obstruction groups. The Fisher exact test was used to compare groups for differences in sex, donor type, method for ureter patency, and rates. P values < .05 were chosen to be statistically significant. A Kaplan-Meier survival analysis was done for graft survival between each group, and censored group graft survival analyses with intervals were performed. Graft survival rates were compared with log-rank (Mantel-Cox) tests between early and late obstruction groups and also between those with successful and unsuccessful interventional treatment.
Among 27 patients included in the study, 20 ureteral obstructions (7 early and 13 late) were detected (Table 1). One patient had both urinary leak and early obstruction and was included in both groups. The level of obstruction was distal ureter in 13 patients, ureteropelvic junction in 4 patients, proximal ureter in 1 patient, and multiple levels in 2 patients. The mean age for the early obstruction group was 25 years (range, 8-49 y), and the mean age for the late obstruction group was 37 years (range, 13-64 y), with mean age significantly higher for the late obstruction group (P = .045). The time of diagnosis was 16 to 88 days (mean of 47 d) in the early obstruction group and 4 to 168 months (mean of 31 mo) in the late obstruction group. Mean preprocedural creatinine level was 2.4 mg/dL (range, 0.5-4.59 mg/dL), and mean postprocedural creatinine was 1.6 mg/dL (range, 0.8-4.7 mg/dL) after PCN (P = .001). There were no major procedure-related complications. There were 2 minor complications of self-dissolving hematuria (6.6%), which were grade B according to the quality improvement guidelines of the Society of Interventional Radiology. Complications involving hematuria were seen only after PCN, and no major complications were detected after further intervention procedures. No allograft loss was seen as a consequence of procedures. No immediate or 30-day mortality was seen after any interventional procedure.
In the early obstruction group (n = 7), the technical success rate of PCN, DJS placement, and BD was 100%. Ureter patency was achieved in 3 patients (43%) without any need for an additional procedure. In 4 patients (57%), after removal of the DJS, obstruction had recurred (range, 1 d to 2 mo) and PCN treatment was repeated (Table 2 and Figure 1). One patient was treated with only PCN as antegrade pyelography revealed a clot in the pelvic junction due to DJS removal, which was extracted with endoscopy. This patient showed no signs of recurrence in his previous obstruction site (patient 5 in Table 2). The remaining patients had recurrences in their obstruction site, and interventional steps of BD were repeated in 3 patients, with CBD used in 2 of these patients. In 1 patient, TUS was placed after CBD. In long-term follow-up (mean follow-up of 122 mo; range 6-184 mo), ureter patency was achieved in all patients in the early obstruction group (100%). Other than 1 patient who lost the graft because of allograft nephropathy, grafts in all other patients were still functioning at time of writing of this study.
In the late obstruction group (n = 13 patients), the technical success rate of PCN was 100%. In 3 patients, no decline in creatinine levels was observed after PCN and biopsy revealed rejection and no further procedure was done. In 1 patient who had an additional VACTERL association (vertebral defects, anal atresia, cardiac defects, tracheoesophageal fistula, renal anomalies, and limb abnormalities) and bladder anomaly, after PCN, the patient underwent bladder revision and ureteroneocystostomy. In the remaining 9 patients, DJSs were placed in 8 patients. In 1 patient, crossing the level of obstruction was unsuccessful, and ureter patency was achieved with ureteroneocystostomy (patient 1 in Table 3). The technical success rate of DJS insertion was 88% (8/9 patients) in the late obstruction group. Ureter patency was achieved in 3 of 9 patients (33%) without the need for an additional procedure. In 5 patients after removal of the DJS, obstruction findings recurred (range, 1 d to 55 mo) (Table 3 and Figure 1). Two of these patients were treated with surgery after PCN, and 3 of these patients required interventional repeated steps of BD, CBD, and DJS. Among these 3 patients, 1 patient had no response to the second intervention of BD and CBD and surgery was performed. The remaining 2 patients had 3 and 4 more recurrences, with 1 having unsuccessful surgical treatment. These 2 patients were followed until rejection with DJS exchange (patients 6 and 7 in Table 3). Two patients were lost to follow-up at 93 and 168 months but with functioning grafts at that time. One patient died of unrelated cause (death was at 93 months), although with a functioning graft at time of death. Two patients had graft loss due to chronic allograft nephropathy at 96 and 120 months. The remaining 4 patients had functioning grafts at time of writing of this study (range, 113-185 mo). Mean follow-up was 132 months (range, 93-185 mo) in the late obstruction group.
Among the 27 patients included in the study, 8 patients had urinary leaks. One patient had an additional fistula extending to the skin. One patient had additional obstruction findings in the distal end. The localization of the leak was the anastomosis site in 7 patients and the renal pelvis in 1 patient. The mean age of patients in the urinary leak group was 38 years (range, 18-60 y). The time of diagnosis was 5 to 90 days posttransplant (mean of 34.2 d).
Percutaneous nephrostomy was the first treatment step in 7 patients, with the other patient having a nephroureteral catheter. The rate of technical success was 100%. Percutaneous drainage catheters were also placed in addition to PCN in 5 patients for large urinomas. In 5 patients, PCN was followed with DJS placement within 5 to 10 days, with technical success rate of 100%. Two patients underwent direct surgical correction after PCN because of leak size, 1 patient underwent surgery because no decrease was observed in the leak amount after nephroureteral stent (NUS) placement, and 1 patient underwent surgery because of recurrence of the leak after removal of DJS. The mean disappearance time for urinary leakage was 29 days (range, 17-30 d). In the patient with additional ureter stricture, the leak resolved but obstructive findings recurred, which were treated with CBD and TUS. Overall, ureter patency was achieved by interventional methods in 4 patients (50%) in long-term follow-up (Table 4 and Figure 2). One patient was lost to follow-up at 24 months, but with a functioning graft at that time. Two patients died due to unrelated causes (time of death was at 3 and 120 months) but had functioning grafts at time of death. Graft loss was seen in 2 patients: 1 from sepsis and 1 from chronic allograft nephropathy (range, 7-84 mo). No procedural-related graft loss was seen. The remaining 3 patients had functioning grafts at time of writing of this study (range, 120-178 mo). The mean follow-up was 86.5 months (range, 3-178 mo).
Outcomes in comparisons of groups
Kaplan-Meier-censored graft survival analyses in all study groups (early, late obstruction, and urinary leak) at 1, 5, and 10 years showed graft survival rates of 89%, 89%, and 73.9%, respectively. The success rates of long-term ureter patency achieved with percutaneous BD and DJS were 100%, 33%, and 50% for early obstruction, late obstruction, and urinary leak groups, respectively (P = .018). However, the method of treatment (interventional vs surgery/failed treatment) showed no significant difference among graft survival rates among groups (P = .142) (Figure 3). When early and late obstruction groups were analyzed, censored group graft survival rates at 1 and 5 years were 85.7% and 84.6%, respectively. Decreased 8- and 10-year graft survival rates were shown in the late obstruction group of 75.2% and 60.2%, respectively, whereas the rate remained the same for the early group at 85.7%. However, when we compared graft survival rates between the early and late obstruction groups, there was no significant difference (P = .394) (Figure 4).
Our results showed that percutaneous methods were safe and effective in the treatment of posttransplant ureteral obstruction and leaks, with no major complication, mortality, or periprocedural graft loss. In addition, although the technical success rates were better in those with early obstructions with long-term ureter patency rates, graft survival rates were not significantly different from those in the late obstruction group. Among similar studies, our study had the longest follow-up period with a mean follow-up of 122 months (range, 6-184 mo) in the early obstruction group, 132 months (range, 93-185 mo) in the late obstruction group, and 86.5 months (range, 3-178 mo) in the urinary leak group.
Success rates of percutaneous management of early and late obstruction have been reported to range from 58% to 100% and 16% to 66%.4-6 During early follow-up, Aytekin and associates7 achieved 100% ureter patency in 9 patients in their early obstruction group and in 10 patients in their late obstruction group. However, recurrence was seen in 4 patients in each group (44%) in long-term follow-up, which was 41 months in the early obstruction group and 23.2 months in the late obstruction group. These recurrences were treated with BD and CBD, and patients who did not respond were treated with metallic stent insertion. In our study, the higher success rate seen among patients with early obstruction versus those with late obstruction was consistent with this report and also with other investigations.5,7 The difference is usually explained as ischemic fibrosis being the main reason for late obstruction, whereas early obstruction tends to be due to mechanical causes, including kinks, edema, clots, or restrictive submucosal tunnel.8
Various results have been published with regard to factors the affect success rates, such as onset of obstruction, length of stricture, and different interventional methods (PCN, DJS, BD, CBD, and metallic stents).4-9 However, no consensus has been reached on the method of choice and definitive indications for management protocols. Uflacker and colleagues10 compared graft survival after NUS placement with and without BD in 67 patients. They reported overall graft survival rates at 1, 3, and 6 years of 92%, 79%, and 59%, respectively. The investigators found no significant difference between patients who received NUS only and those who received NUS and BD with regard to survival from the time of tube removal (P = .567). When the investigators divided their cohort of strictures into early (< 3 mo) and late (> 3 mo) groups, there was a statistically significant difference between the groups, with median graft survival of 107 months for the early group and 65.3 months for the late group (P = .032). In our study cohort, graft survival rates at 1, 5, and 10 years were 89%, 88%, and 73.9%, respectively, and differences in graft survival rates between the early and late obstruction groups were not statistically significant, although ureter patency rate was higher in the early obstruction group.
Ooms and associates11 reported a technical success rate of 86% of antegrade BD in 50 patients with ureteric obstructions. However, clinical success was achieved in 47% of their patients. Although they reported success rates without classifying the obstructions according to time of onset, in their risk factor analysis, time of onset was not found to be a risk factor that could influence treatment outcome. The length of the stricture and the type or size of the balloon that was used were also not found to be risk factors. The investigators did not report a significant difference in graft survival rates between successful and unsuccessful procedures of BD; however, the study had a shorter follow-up time than our study.
Simsek and associates12 compared PCN following antegrade stenting with open surgery and also found no statistically significant difference in graft survival. They classified their cohort according to the length of stenosis and mentioned a benefit in favor of open surgery in type 2 (partial/dense fibrosis effecting ≤ 50% of the ureter) obstructions in terms of decreased reintervention rates, although this was not statistically significant in terms of graft survival outcome. Our graft survival results are consistent with this study, although our study design was not a direct comparison between percutaneous methods and surgical methods.
Kumar and associates13 studied graft survival in 52 patients with percutaneous ureteroplasty versus a matched control group of transplant recipients with no history of ureteric stenosis. They found no significant difference in 10-year graft survival and timing onset of stenosis. With outcomes not significantly worse than those for the control group, they pointed out the minimal invasive nature of percutaneous methods with lower morbidity and similar long-term outcomes. Our study results and the technique of percutaneous methods that we use (BD followed by DJS placement) are similar to findings in their study. Also parallel with their study, minor complications occurred only after PCN placement and no major complication or 30-day mortality or graft loss was seen after BD and DJS insertion.
Published data on use of cutting balloons in ureteral strictures are limited in renal transplant patients. Published outcomes of these techniques in renal transplant recipients so far include only small case series and retrospective cohorts.7,14-16 In our present study, CBD was performed in 4 patients (2 in the early group, 2 in the late group) with recurrences. In the early obstruction group, clinical success was achieved with follow-up of 84 and 156 months. However, in the late obstruction group, a stricture in 1 patient did not respond to CBD and the patient underwent surgical correction. In the second patient, although an initial response was seen, a third recurrence occurred after 8 months, and the patient underwent surgery. However, surgical correction was also not successful in this patient. More multicenter prospective studies are needed to show the efficacy and superiority of CBD versus other standard procedures.
In our study, 4 patients underwent a novel attempt of ureteral obstruction treatment with TUS placement. Tandem ureteral stents were first used in malignant ureteral obstructions and have shown good outcomes and stent patency rates of up to 4 to 6 months in retrospective studies.17-20 Liu and colleagues21 showed no difference in survival rates but significantly longer stent patency rates in patients with TUS in their prospective study of 104 malignant ureteral obstructions. The investigators’ goal in inserting TUS was to preserve lumen patency for a longer period of time after BD as an extra lumen is created between them versus that with a single DJS, providing urine to flow. Because, on average, a single DJS is usually removed at 3 months due to infection and/or obstruction of the ureteral stent, TUS may extend this time up to 6 months. So far, there are only 2 studies with TUS placement in transplant ureters. Miyaoka and associates22 placed TUS after attempts to achieve ureter patency with BD and/or DJS were not successful in ureteric obstructions after renal transplant. They reported a success rate of 58% in 19 patients with resistant stenosis after renal transplant. In their study, which had a follow-up of 62.8 months, 7 patients (37%) remained patent after removal of TUS. The mean time that the tandem stents were in place was 49 ± 48 weeks (95% confidence interval, 20-78 wk) in successfully treated patients. On the other hand, Kriegshauser and associates23 used TUS in 18 patients as primary treatment in the management of transplant ureteric strictures. They reported a success rate of 83% (15/18 patients) without surgery with a mean follow-up of 264 days (range, 92-757 d) after stent removal. The time range of the stents in place was 33 to 304 days in their study. They suggested more intensive monitoring for urinary tract infections in patients with TUSs, as urinary tract infections occurred in 14 of 18 patients (78%).
In our study, technical success was achieved in all 4 patients with TUS. Two of these patients were in the early obstruction group, with TUS inserted after BD. One showed primary ureter patency at 172-month follow-up, and the other patient showed a secondary ureter patency at 157-month follow-up without any loss of graft function. However, 2 patients in the late obstruction group showed recurrences on long-term follow-up, although the initial results were successful. Of these patients, 1 ureter was corrected surgically. The other patient had unsuccessful surgical revision; this patient was followed up with TUS exchanged every 6 months until rejection (follow-up of 96 mo). Graft loss was avoided in this patient with novel interventional techniques when surgery failed. It is also believed that TUS is better at resistance of obstruction by providing space in between the 2 stents that is difficult to compress.19 Although the present results are not powered statistically to back up this theory and more prospective multicenter studies are needed, in resistant cases of ureteric strictures, TUS may be a method of choice, especially in patients with unsuccessful surgery and who require lifelong ureteral stents.
Urine leaks generally present in the immediate or early posttransplant period. Technical complications during transplant in the immediate period and insufficient blood supply to the ureter causing ischemia and necrosis in the early period are the main causes, and most leaks occur at the distal portion of the ureter.24-26 Alcaraz and colleagues27 suggested 60% of patients with urine leaks developed in 3 days would benefit from only PCN treatment. Surgical correction is advocated in cases of large volume leaks and proximal location.28 In our cohort, 2 patients had surgical correction after PCN due to volume and extension of the leak. In the remaining patients, PCN was followed with DJS placement. Among these patients, ureter patency rate was 80% (5/7 patients) in long-term follow-up by interventional methods. The success rates reported with interventional methods other than PCN have ranged from 36% to 100%.5,7,28-30 So far, it should be kept in mind that there is no study showing superiority of surgical correction to interventional methods, although the literature is based on evidence from retrospective studies.
In management of ureteral complications, favorable surgical outcomes and endourologic procedures have been documented.31,32 However, no superiority of one to another has been shown. Perioperative graft loss is a risk with open surgery,33 and perioperative morbidity rates are reported as 20%.34 However, no graft loss or 30-day mortality has been reported in the literature with percutaneous methods of BD, CBD, or DJS placement. Patel and colleagues35 suggested that current practice parameters of technical and clinical success rates and complications of PCN are mostly for native kidneys, and they suggested that focused studies on technical and clinical success rates, outcomes, and comui1plications after transplant nephrostomy could identify practice parameters. In their study, 30-day graft and mortality rates were 0% among 73 PCN procedures. Their overall complication rate was 6.8%, and rate of major complications was 1.4%. Similar to this study, our investigation showed a minor complication rate of 6.6% among 27 patients, with no major complication. Immediate and 30-day graft mortality rates were 0%.
The limitations of our study include its retrospective design and the relatively small number of patients. Although no differences were found in graft survival rates, the influence of medical management regimens on graft survival was lacking in our study, which may have changed during the long-term follow-up. In addition, with our study design, we could not definitively conclude superiority between methods, although graft survival rates were similar to those shown in the literature.
Percutaneous management of ureteral obstructions and leaks are safe and effective and should be considered as first-line treatment because of their less invasive nature and lower complication and morbidity rates. The use of CBD and TUS may increase rates of clinical success.
Volume : 19
Issue : 8
Pages : 788 - 798
DOI : 10.6002/ect.2019.0422
From the 1Department of Radiology, the 2Department of Nephrology, the
3Department of Biostatistics, and the 4Department of Urology, Hacettepe
University Faculty of Medicine, Ankara, Turkey
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare.
Corresponding author: Fatma Gonca Eldem, Hacettepe University Faculty of Medicine Department of Radiology, HUTF Radyoloji Anabilim Dalı, 06100, Sihhiye, Ankara, Turkey
Phone: +90 312 305 1880
Table 1. Patient Demographics and Management Characteristics Between Early and Late Obstruction and Urinary Leak Groups
Table 2. Patient Characteristics and Outcomes in Early Obstruction Group
Table 3. Patient and Treatment Characteristics and Outcomes in Late Obstruction Group
Table 4. Patient and Treatment Characteristics and Outcomes in the Urinary Leak Group
Figure 1. Flow Chart of Management of Patients in the Obstruction Groups
Figure 2. Flow Chart of Management of Patients in Urinary Leak Group
Figure 3. Log-rank (Mantel-Cox) Test Comparison of Graft Survival Rates Between Successful and Failed (or Surgery) Interventional Procedures
Figure 4. Log-Rank (Mantel-Cox) Test Comparison of Graft Survival Rates Between Early and Late Obstruction Groups