The role of prophylactic versus selective ureteric stenting in kidney transplant patients: a retrospective review Objective—To evaluate the role of prophylactic versus selective ureteric stenting in the development of postoperative ureterovesical complications in kidney transplant recipients. Methods—Records of 614 transplant patients seen from January 2006 to May 2011 were retrospectively reviewed. The primary outcome was the rate of ureterovesical complications, defined as the development of ureteric obstruction or a ureterovesical anastomotic leak. The secondary outcomes were the rate of urinary tract infections and forgotten stents. Using a χ2 test, we compared the primary and secondary outcomes across the selective and prophylactic cohorts. Logistic regression was used to compare the 2 cohorts while adjusting for potential confounders. Results—The selective and prophylactic cohorts consisted of 258 and 330 patients, respectively. Unadjusted analysis showed that the prophylactic group had a significantly lower rate of ureterovesical complications than did the selective group (2.12% vs 6.20%; odds ratio, 0.33; P = .01). After adjustment for differences in sex and donor type, the prophylactic group still had a lower risk for ureterovesical complications (odds ratio, 0.30; P = .009). Rates of urinary tract infections and forgotten stents did not differ significantly between the 2 groups. Conclusions—Prophylactic stenting is associated with a significantly lower rate of ureterovesical complications than is selective stenting. (Progress in Transplantation. 2014;24:322-327) ©2014 NATCO, The Organization for Transplant Professionals doi: http://dx.doi.org/10.7182/pit2014422
U
rologic complications after kidney transplant most commonly originate from the ureterovesical anastomosis and often occur soon after transplant,1 contributing to patients’ morbidity, graft loss, mortality,2,3 and prolonged hospital stay. Ureterovesical complications after kidney transplant manifest predominantly as a urinary leak or ureteral obstruction, both of which have been successfully treated with placement of a ureteric stent. This success has led to the notion of prophylactic stenting in all patients in an effort to prevent these complications. Although numerous studies, 1,4-10 along with 2 metaanalyses,11,12 have compared the use of prophylactic stents with no stents in preventing ureterovesical complications, very few studies have examined the use of prophylactic versus selective stenting.13,14 Considering that ureteral stents might be a source of morbidity
322
Michael Ordon, MD, MSc, FRCSC, Daniela Ghiculete, MD, MSc, Robert Stewart, MD, FRCSC, Kenneth T. Pace, MD, MSC, FRCSC, R. John DʼA. Honey, MD, FRCSC St Michaelʼs Hospital, University of Toronto, Canada Corresponding author: Michael Ordon, MD, MSc, FRCSC, University of Toronto, St Michaelʼs Hospital, 61 Queen St E, Ste 9-103, Toronto, Ontario M5C 2T2, Canada (e-mail: [email protected]) To purchase electronic or print reprints, contact: American Association of Critical-Care Nurses 101 Columbia, Aliso Viejo, CA 92656 Phone (800) 899-1712 (ext 532) or (949) 448-7370 (ext 532) Fax (949) 362-2049 E-mail [email protected]
in the form of urinary tract infection (UTI), stent migration, hematuria, and signs and symptoms in the lower part of the urinary tract, there might be a role for selective stenting to minimize this potential morbidity, while still maintaining the benefits of stents in certain cases. The objective of this study was to determine whether prophylactic ureteral stenting at the time of kidney transplant reduces the rate of ureterovesical complications in the postoperative period compared with selective stenting. Methods Patients A retrospective cohort study was performed by completing a chart review of all patients who had undergone kidney transplant at St Michael’s Hospital, Toronto, Canada, between January 2006 and May 2011.
Progress in Transplantation, Vol 24, No. 4, December 2014
The role of prophylactic versus selective ureteric stenting Demographic and surgical data including patients’ age, sex, cause of renal failure, living or deceased donor, surgeon, and use of ureteric stent were recorded for all patients. All patients who underwent kidney transplant between January 2006 and May 2011 and had at least 3 months of postoperative follow-up at St Michael’s Hospital were included in the study. Patients whose graft did not survive to 3 months postoperatively were excluded from the study cohort, as they did not have a functioning graft for long enough for the outcome of interest to develop. This study was conducted with approval of the research ethics board at St Michael’s Hospital. Exposure/Study Cohorts All patients who underwent a kidney transplant between January 1, 2006, and September 1, 2008, constituted the selective stent cohort. In this group of patients, per the institutional protocol, a ureteric stent was placed at the time of kidney transplant at the discretion of the staff surgeon. Stents were placed primarily in cases where the transplant showed evidence of ureteral devascularization from retrieval or the recipient had a small-caliber thin-walled ureter, a thickwalled bladder, a neurogenic bladder, a technically challenging ureteral anastomosis, or scarring because of a prior transplant on the ipsilateral side. Patients who underwent kidney transplant between September 2, 2008, and May 31, 2011, constituted the prophylactic stent cohort. Following a change in institution protocol that occurred at this time, all patients in this cohort received a prophylactic ureteric stent at the time of transplant. In all patients, the ureterovesical anastomosis was done by using an extravesical technique. Additionally, in both cohorts stents were removed cystoscopically, with the patient under local anesthetic, at approximately 4 weeks after transplant. Outcomes The primary outcome was the development of a ureterovesical complication within 3 months postoperatively. Ureterovesical complication was defined as either ureteral obstruction or ureterovesical anastomotic leak. The diagnoses of ureteral obstruction and ureterovesical anastomotic leak were determined by a review of imaging studies, results of diagnostic laboratory tests, and clinical notes. If known, the cause and location of the ureteral obstruction were recorded. The secondary outcomes were the development of a UTI postoperatively and forgotten stents. A UTI was defined as a urine culture that showed growth of microorganisms (colony count ≥ 1 × 105 colonyforming units/mL) within the first 6 weeks postoperatively. A stent was defined as forgotten if it remained
Progress in Transplantation, Vol 24, No. 4, December 2014
Transplants reviewed (N = 614) Patients excluded (n = 26)
Study cohort (n = 588)
Selective cohort (n = 258)
Deceased (n = 5) Failed graft (n = 21) • Vascular (n = 9) • Coagulopathy/hemodynamic (n = 5) • Deep venous thrombosis (n = 2) • Other (n = 5)
Prophylactic cohort (n = 330)
Figure Selection of patients for inclusion in the study.
in place for longer than 3 months postoperatively, as routinely stents are removed 4 weeks after transplant. Statistical Analysis Demographic and surgical data were compared across the 2 cohorts by using the Student t test for continuous variables and the χ2 test for categorical variables. Both the primary and secondary outcomes were compared across the 2 cohorts by using the χ2 test (or the Fisher exact test when applicable). Subset analysis was performed to assess each component of the primary outcome independently (ie, ureteral obstruction and ureterovesical anastomotic leak) and to compare the primary outcome across the deceased donor and living donor cohorts separately. Logistic regression analysis was used to compare the primary outcome across the 2 cohorts while controlling for differences in sex and donor type. All data analysis was done by using SAS software version 9.2 (SAS institute). A P value of .05 or less was considered to indicate statistical significance. Sample Size Calculation The χ2 method was used to estimate the sample size. A sample size of 334 patients (167 per group) was needed to provide a 2-sample, 2-sided test of proportions with 80% power and an α of .05 to detect a 6% absolute difference between groups given an estimated 1% event rate12 in the prophylactic stent group. Results Of the 614 patients retrospectively reviewed, 26 patients (12 from the prophylactic stent cohort and 14 from the selective stent cohort) were excluded because their renal allograft did not survive 3 months. Nineteen of these patients required a transplant nephrectomy, and 5 patients died (Figure). None of the excluded
323
Ordon et al Table 1 Patient demographics and surgical characteristics
Characteristic
Selective cohort (n = 258)
Prophylactic cohort (n = 330)
Pa .08
50.7 (12.4)
52.6 (13.3)
Sex, % Male Female
65.9 34.1
57.3 42.7
Allograft, % Live donor Deceased donor
46.5 53.5
37.3 62.7
Stent, % Yes No
26.4 73.6
99.7 0.3
Surgeon, % 1 2 3 4
47.7 34.5 16.7 1.2
42.4 39.1 17.3 1.2
Age, mean (SD), y
a P < .05
.03
.02
—
.63
statistically significant. Dash indicates not applicable.
patients had a ureterovesical complication before nephrectomy or death. The remaining 588 patients comprised our study cohort. The prophylactic stent cohort consisted of 330 patients and the selective stent cohort had 258 patients. The 2 groups were similar in mean age and performing surgeon, but differed with respect to sex and donor allograft type (Table 1). In the selective stent group, 26.4% of patients had a stent placed, compared with 99.7% in the prophylactic stent group. For the primary outcome, unadjusted analysis showed that the proportion of ureterovesical complications in the prophylactic stent group (2.12%) was significantly lower than that in the selective stent group (6.20%; odds ratio, 0.33; 95% CI, 0.13-0.81, P = .01). Considering that the 2 cohorts differed significantly with respect to sex and donor allograft type, multivariable logistic regression was performed to adjust for potential confounding from either of these 2 covariates on the primary outcome. After differences in sex and donor allograft type across the 2 groups were controlled for with logistic regression, the prophylactic stent group was still associated with a significantly lower risk for ureterovesical complication (odds ratio, 0.30; 95% CI, 0.12-0.74; P = .009). In this multivariable analysis, sex was not a significant predictor of ureterovesical complication, but donor allograft type was, specifically grafts from deceased donors were associated with a greater risk of ureterovesical complication (odds ratio, 2.96; 95% CI, 1.07-8.14; P = .04) when stent group (prophylactic vs selective) was controlled for. Examining the 2 components of ureterovesical complication separately revealed that only ureteral obstruction differed significantly between the 2 groups. The rate of ureteral obstruction was 1.21% in the prophylactic group and 4.26% in the selective
324
group (P = .03). Conversely, the rate of ureterovesical leak was 0.91% in the prophylactic cohort and 3.1% in the selective cohort (P = .07). Subset analysis examining the effect of prophylactic versus selective stent use only in the patients with a live donor allograft showed no significant difference (P = .21) in ureterovesical complication rate between the prophylactic and selective cohorts. Conversely, in the patients with a deceased donor allograft, prophylactic stenting showed a significant benefit over selective stenting with respect to the development of the primary outcome (odds ratio, 0.31; 95% CI, 0.11-0.86; P = .02). Additional subset analysis of the selective cohort revealed no significant difference in the rate of ureterovesical complications between patients who received a stent and patients who did not receive a stent (7.35% vs 5.79%, P = .78). The definitive treatments for all cases of ureteral obstruction in both cohorts are listed in Table 2. In all cases, obstruction was initially treated with percutaneous nephrostomy or nephroureterostomy tube placement. The exception to this was 2 cases where obstruction was due to a large lymphocele, which was treated successfully with percutaneous drainage of the lymphocele. Four patients required an intervention beyond temporary nephrostomy or nephroureterostomy tube insertion for definitive management of their ureteral obstruction (Table 2). Additionally, the definitive treatments for all cases of ureterovesical anastomotic leak are also detailed in Table 2. Eight cases required further definitive intervention beyond temporary catheter drainage (Table 2). For our secondary outcomes, analysis with the χ2 test revealed that there was no significant difference in
Progress in Transplantation, Vol 24, No. 4, December 2014
The role of prophylactic versus selective ureteric stenting Table 2 Definitive treatments for ureterovesical complications Treatment of ureteral obstruction
Table 3 Outcome of patients with a forgotten stent
Treatment of ureterovesical leaks
Percutaneous nephrostomy or nephroureterostomy tube insertion (only)
9 Foley catheter (only)
3
Percutaneous balloon dilatation
2 Percutaneous drain (+catheter)
6
Open repair
2 Open repair
2
the proportion of UTIs (20.9% vs 16.7%, P = .19) or forgotten stents (4.2% vs 3.9%, P = .82) between the selective and prophylactic stent cohorts, respectively. In Table 3, the management for all patients in both cohorts who had a forgotten stent is detailed. Discussion In this retrospective cohort study, we examined the impact of prophylactic stenting as compared with selective stenting at the time of kidney transplant on the development of postoperative ureterovesical complications. We also assessed the effect of both stenting practices on the rates of UTI and forgotten stents postoperatively. Our study revealed significantly fewer ureterovesical complications in the prophylactic stent group than in the selective stent group (odds ratio, 0.33; P = .01). After differences in sex and allograft type across the 2 groups had been adjusted for with logistic regression analysis, the benefit was even more pronounced in the prophylactic stent cohort (odds ratio, 0.30; P = .009). Not surprisingly, receiving a deceased donor allograft was a strong predictor of ureterovesical complication in the regression analysis (odds ratio, 2.96; P = .04), when study group was controlled for. Subset analysis in which the effect of prophylactic versus selective stenting was assessed separately in the patients who received a living donor vs a deceased donor allograft, benefit of prophylactic stent placement was apparent only in the recipients of deceased donor allografts. The evidence in the literature is conflicting with regards to this issue, with trials both showing benefit and no benefit of prophylactic stent placement in patients receiving a living donor allograft.6,7 One might expect superior integrity of the microvascular supply to the ureter in living donor allografts, as such allografts have shorter cold ischemic times, more meticulous dissection with immediate cold perfusion during procurement, and the donors are typically very healthy. Conversely, with the chronic shortage of available organs from deceased donors, the acceptance criteria for donors have expanded to include older patients and patients with a history of
Progress in Transplantation, Vol 24, No. 4, December 2014
Selective cohort (n = 10)
Prophylactic cohort (n = 14)
No additional intervention (cystoscopic removal)
8
10
Shockwave lithotripsy
1
0
Percutaneous procedure
0
2
Combined ureteroscopy and percutaneous procedure
1
0
Undetermined
0
2
Outcome
hypertension, diabetes, and smoking, all of which might result in more compromised microvasculature. In addition, the ureterovesical anastomosis can be more challenging with a thick-walled, small-capacity bladder, more commonly seen in recipients of deceased donor allografts, who have often been treated with dialysis for many years while waiting for a transplant. Recipients of living donor allografts are healthier, and their transplants may be preemptive. Accordingly, these patients often have normal bladders. Importantly, however, our study might have been underpowered to detect a difference in ureterovesical complications in the recipients of living donor allografts only, as this group constituted a smaller subset of our entire study sample. Additional subset analysis evaluating the rate of ureterovesical complications within the selective group revealed no significant difference between the patients who received a stent and patients who did not. One might have expected the patients with no stent to have a higher complication rate because of the absence of a stent. However, the patients with a stent most likely had a greater predisposition to a complication developing, as the surgeons would typically place stents in more complex cases at risk for complications. Again, our study might have been underpowered to detect a difference in ureterovesical complications in the selective cohort only, as this constituted a smaller subset of our entire study sample. Despite the increased use of stents in the prophylactic group, we did not see an increase in either forgotten stents or UTIs after transplant. Our finding with respect to UTIs is contrary to what was reported in the meta-analysis by Wilson et al.12 In their meta-analysis, they identified a significant increase in UTIs associated with stents (risk ratio, 1.49; 95% CI, 1.04-2.15, P = .03). However, heterogeneity was significant (χ2 = 15.32, P = .02, I2 = 60.8%) in their pooled results for this outcome. Subsequent subgroup analysis revealed that the risk was associated with the antibiotic regime. Specifically, studies detailing prophylaxis with longterm cotrimoxazole had an equivalent risk ratio for
325
Ordon et al infection between patients with a stent and patients without a stent (risk ratio, 0.97; 95% CI, 0.71-1.33) compared with other regimes. Accordingly, at our institution, all transplant patients are maintained on longterm therapy with cotrimoxazole for a minimum of 1 year, and this fact might explain our findings. Nevertheless, the results of our secondary outcomes must be interpreted with caution, because our study was not powered to detect a difference in these outcomes. Our forgotten stent rate was higher than expected in both groups, highlighting the importance of a coordinated system to ensure that all patients return for stent removal. Despite some lengthy indwelling times for the forgotten stents, most patients (75%) did not require any additional intervention for stent removal beyond a cystoscopy. The absence of encrustation and stone formation due to the forgotten stent is in keeping with the previously reported low rates of allograft lithiasis.15-17 As previously mentioned, the use of prophylactic stents compared with no stent following kidney transplant has been well studied. In a meta-analysis published in 2004, Mangus and Haag11 reviewed 49 articles published between 1973 and 2002 evaluating the use of ureteric stents at the time of kidney transplant. The meta-analysis included 5 randomized controlled trials and 44 case series. In the controlled trials group, urologic complications occurred in 6 of 407 patients with a stent (1.5%) and 35 of 389 patients without a stent (9.0%; odds ratio, 0.24; 95% CI, 0.10-0.57; P
U
rologic complications after kidney transplant most commonly originate from the ureterovesical anastomosis and often occur soon after transplant,1 contributing to patients’ morbidity, graft loss, mortality,2,3 and prolonged hospital stay. Ureterovesical complications after kidney transplant manifest predominantly as a urinary leak or ureteral obstruction, both of which have been successfully treated with placement of a ureteric stent. This success has led to the notion of prophylactic stenting in all patients in an effort to prevent these complications. Although numerous studies, 1,4-10 along with 2 metaanalyses,11,12 have compared the use of prophylactic stents with no stents in preventing ureterovesical complications, very few studies have examined the use of prophylactic versus selective stenting.13,14 Considering that ureteral stents might be a source of morbidity
322
Michael Ordon, MD, MSc, FRCSC, Daniela Ghiculete, MD, MSc, Robert Stewart, MD, FRCSC, Kenneth T. Pace, MD, MSC, FRCSC, R. John DʼA. Honey, MD, FRCSC St Michaelʼs Hospital, University of Toronto, Canada Corresponding author: Michael Ordon, MD, MSc, FRCSC, University of Toronto, St Michaelʼs Hospital, 61 Queen St E, Ste 9-103, Toronto, Ontario M5C 2T2, Canada (e-mail: [email protected]) To purchase electronic or print reprints, contact: American Association of Critical-Care Nurses 101 Columbia, Aliso Viejo, CA 92656 Phone (800) 899-1712 (ext 532) or (949) 448-7370 (ext 532) Fax (949) 362-2049 E-mail [email protected]
in the form of urinary tract infection (UTI), stent migration, hematuria, and signs and symptoms in the lower part of the urinary tract, there might be a role for selective stenting to minimize this potential morbidity, while still maintaining the benefits of stents in certain cases. The objective of this study was to determine whether prophylactic ureteral stenting at the time of kidney transplant reduces the rate of ureterovesical complications in the postoperative period compared with selective stenting. Methods Patients A retrospective cohort study was performed by completing a chart review of all patients who had undergone kidney transplant at St Michael’s Hospital, Toronto, Canada, between January 2006 and May 2011.
Progress in Transplantation, Vol 24, No. 4, December 2014
The role of prophylactic versus selective ureteric stenting Demographic and surgical data including patients’ age, sex, cause of renal failure, living or deceased donor, surgeon, and use of ureteric stent were recorded for all patients. All patients who underwent kidney transplant between January 2006 and May 2011 and had at least 3 months of postoperative follow-up at St Michael’s Hospital were included in the study. Patients whose graft did not survive to 3 months postoperatively were excluded from the study cohort, as they did not have a functioning graft for long enough for the outcome of interest to develop. This study was conducted with approval of the research ethics board at St Michael’s Hospital. Exposure/Study Cohorts All patients who underwent a kidney transplant between January 1, 2006, and September 1, 2008, constituted the selective stent cohort. In this group of patients, per the institutional protocol, a ureteric stent was placed at the time of kidney transplant at the discretion of the staff surgeon. Stents were placed primarily in cases where the transplant showed evidence of ureteral devascularization from retrieval or the recipient had a small-caliber thin-walled ureter, a thickwalled bladder, a neurogenic bladder, a technically challenging ureteral anastomosis, or scarring because of a prior transplant on the ipsilateral side. Patients who underwent kidney transplant between September 2, 2008, and May 31, 2011, constituted the prophylactic stent cohort. Following a change in institution protocol that occurred at this time, all patients in this cohort received a prophylactic ureteric stent at the time of transplant. In all patients, the ureterovesical anastomosis was done by using an extravesical technique. Additionally, in both cohorts stents were removed cystoscopically, with the patient under local anesthetic, at approximately 4 weeks after transplant. Outcomes The primary outcome was the development of a ureterovesical complication within 3 months postoperatively. Ureterovesical complication was defined as either ureteral obstruction or ureterovesical anastomotic leak. The diagnoses of ureteral obstruction and ureterovesical anastomotic leak were determined by a review of imaging studies, results of diagnostic laboratory tests, and clinical notes. If known, the cause and location of the ureteral obstruction were recorded. The secondary outcomes were the development of a UTI postoperatively and forgotten stents. A UTI was defined as a urine culture that showed growth of microorganisms (colony count ≥ 1 × 105 colonyforming units/mL) within the first 6 weeks postoperatively. A stent was defined as forgotten if it remained
Progress in Transplantation, Vol 24, No. 4, December 2014
Transplants reviewed (N = 614) Patients excluded (n = 26)
Study cohort (n = 588)
Selective cohort (n = 258)
Deceased (n = 5) Failed graft (n = 21) • Vascular (n = 9) • Coagulopathy/hemodynamic (n = 5) • Deep venous thrombosis (n = 2) • Other (n = 5)
Prophylactic cohort (n = 330)
Figure Selection of patients for inclusion in the study.
in place for longer than 3 months postoperatively, as routinely stents are removed 4 weeks after transplant. Statistical Analysis Demographic and surgical data were compared across the 2 cohorts by using the Student t test for continuous variables and the χ2 test for categorical variables. Both the primary and secondary outcomes were compared across the 2 cohorts by using the χ2 test (or the Fisher exact test when applicable). Subset analysis was performed to assess each component of the primary outcome independently (ie, ureteral obstruction and ureterovesical anastomotic leak) and to compare the primary outcome across the deceased donor and living donor cohorts separately. Logistic regression analysis was used to compare the primary outcome across the 2 cohorts while controlling for differences in sex and donor type. All data analysis was done by using SAS software version 9.2 (SAS institute). A P value of .05 or less was considered to indicate statistical significance. Sample Size Calculation The χ2 method was used to estimate the sample size. A sample size of 334 patients (167 per group) was needed to provide a 2-sample, 2-sided test of proportions with 80% power and an α of .05 to detect a 6% absolute difference between groups given an estimated 1% event rate12 in the prophylactic stent group. Results Of the 614 patients retrospectively reviewed, 26 patients (12 from the prophylactic stent cohort and 14 from the selective stent cohort) were excluded because their renal allograft did not survive 3 months. Nineteen of these patients required a transplant nephrectomy, and 5 patients died (Figure). None of the excluded
323
Ordon et al Table 1 Patient demographics and surgical characteristics
Characteristic
Selective cohort (n = 258)
Prophylactic cohort (n = 330)
Pa .08
50.7 (12.4)
52.6 (13.3)
Sex, % Male Female
65.9 34.1
57.3 42.7
Allograft, % Live donor Deceased donor
46.5 53.5
37.3 62.7
Stent, % Yes No
26.4 73.6
99.7 0.3
Surgeon, % 1 2 3 4
47.7 34.5 16.7 1.2
42.4 39.1 17.3 1.2
Age, mean (SD), y
a P < .05
.03
.02
—
.63
statistically significant. Dash indicates not applicable.
patients had a ureterovesical complication before nephrectomy or death. The remaining 588 patients comprised our study cohort. The prophylactic stent cohort consisted of 330 patients and the selective stent cohort had 258 patients. The 2 groups were similar in mean age and performing surgeon, but differed with respect to sex and donor allograft type (Table 1). In the selective stent group, 26.4% of patients had a stent placed, compared with 99.7% in the prophylactic stent group. For the primary outcome, unadjusted analysis showed that the proportion of ureterovesical complications in the prophylactic stent group (2.12%) was significantly lower than that in the selective stent group (6.20%; odds ratio, 0.33; 95% CI, 0.13-0.81, P = .01). Considering that the 2 cohorts differed significantly with respect to sex and donor allograft type, multivariable logistic regression was performed to adjust for potential confounding from either of these 2 covariates on the primary outcome. After differences in sex and donor allograft type across the 2 groups were controlled for with logistic regression, the prophylactic stent group was still associated with a significantly lower risk for ureterovesical complication (odds ratio, 0.30; 95% CI, 0.12-0.74; P = .009). In this multivariable analysis, sex was not a significant predictor of ureterovesical complication, but donor allograft type was, specifically grafts from deceased donors were associated with a greater risk of ureterovesical complication (odds ratio, 2.96; 95% CI, 1.07-8.14; P = .04) when stent group (prophylactic vs selective) was controlled for. Examining the 2 components of ureterovesical complication separately revealed that only ureteral obstruction differed significantly between the 2 groups. The rate of ureteral obstruction was 1.21% in the prophylactic group and 4.26% in the selective
324
group (P = .03). Conversely, the rate of ureterovesical leak was 0.91% in the prophylactic cohort and 3.1% in the selective cohort (P = .07). Subset analysis examining the effect of prophylactic versus selective stent use only in the patients with a live donor allograft showed no significant difference (P = .21) in ureterovesical complication rate between the prophylactic and selective cohorts. Conversely, in the patients with a deceased donor allograft, prophylactic stenting showed a significant benefit over selective stenting with respect to the development of the primary outcome (odds ratio, 0.31; 95% CI, 0.11-0.86; P = .02). Additional subset analysis of the selective cohort revealed no significant difference in the rate of ureterovesical complications between patients who received a stent and patients who did not receive a stent (7.35% vs 5.79%, P = .78). The definitive treatments for all cases of ureteral obstruction in both cohorts are listed in Table 2. In all cases, obstruction was initially treated with percutaneous nephrostomy or nephroureterostomy tube placement. The exception to this was 2 cases where obstruction was due to a large lymphocele, which was treated successfully with percutaneous drainage of the lymphocele. Four patients required an intervention beyond temporary nephrostomy or nephroureterostomy tube insertion for definitive management of their ureteral obstruction (Table 2). Additionally, the definitive treatments for all cases of ureterovesical anastomotic leak are also detailed in Table 2. Eight cases required further definitive intervention beyond temporary catheter drainage (Table 2). For our secondary outcomes, analysis with the χ2 test revealed that there was no significant difference in
Progress in Transplantation, Vol 24, No. 4, December 2014
The role of prophylactic versus selective ureteric stenting Table 2 Definitive treatments for ureterovesical complications Treatment of ureteral obstruction
Table 3 Outcome of patients with a forgotten stent
Treatment of ureterovesical leaks
Percutaneous nephrostomy or nephroureterostomy tube insertion (only)
9 Foley catheter (only)
3
Percutaneous balloon dilatation
2 Percutaneous drain (+catheter)
6
Open repair
2 Open repair
2
the proportion of UTIs (20.9% vs 16.7%, P = .19) or forgotten stents (4.2% vs 3.9%, P = .82) between the selective and prophylactic stent cohorts, respectively. In Table 3, the management for all patients in both cohorts who had a forgotten stent is detailed. Discussion In this retrospective cohort study, we examined the impact of prophylactic stenting as compared with selective stenting at the time of kidney transplant on the development of postoperative ureterovesical complications. We also assessed the effect of both stenting practices on the rates of UTI and forgotten stents postoperatively. Our study revealed significantly fewer ureterovesical complications in the prophylactic stent group than in the selective stent group (odds ratio, 0.33; P = .01). After differences in sex and allograft type across the 2 groups had been adjusted for with logistic regression analysis, the benefit was even more pronounced in the prophylactic stent cohort (odds ratio, 0.30; P = .009). Not surprisingly, receiving a deceased donor allograft was a strong predictor of ureterovesical complication in the regression analysis (odds ratio, 2.96; P = .04), when study group was controlled for. Subset analysis in which the effect of prophylactic versus selective stenting was assessed separately in the patients who received a living donor vs a deceased donor allograft, benefit of prophylactic stent placement was apparent only in the recipients of deceased donor allografts. The evidence in the literature is conflicting with regards to this issue, with trials both showing benefit and no benefit of prophylactic stent placement in patients receiving a living donor allograft.6,7 One might expect superior integrity of the microvascular supply to the ureter in living donor allografts, as such allografts have shorter cold ischemic times, more meticulous dissection with immediate cold perfusion during procurement, and the donors are typically very healthy. Conversely, with the chronic shortage of available organs from deceased donors, the acceptance criteria for donors have expanded to include older patients and patients with a history of
Progress in Transplantation, Vol 24, No. 4, December 2014
Selective cohort (n = 10)
Prophylactic cohort (n = 14)
No additional intervention (cystoscopic removal)
8
10
Shockwave lithotripsy
1
0
Percutaneous procedure
0
2
Combined ureteroscopy and percutaneous procedure
1
0
Undetermined
0
2
Outcome
hypertension, diabetes, and smoking, all of which might result in more compromised microvasculature. In addition, the ureterovesical anastomosis can be more challenging with a thick-walled, small-capacity bladder, more commonly seen in recipients of deceased donor allografts, who have often been treated with dialysis for many years while waiting for a transplant. Recipients of living donor allografts are healthier, and their transplants may be preemptive. Accordingly, these patients often have normal bladders. Importantly, however, our study might have been underpowered to detect a difference in ureterovesical complications in the recipients of living donor allografts only, as this group constituted a smaller subset of our entire study sample. Additional subset analysis evaluating the rate of ureterovesical complications within the selective group revealed no significant difference between the patients who received a stent and patients who did not. One might have expected the patients with no stent to have a higher complication rate because of the absence of a stent. However, the patients with a stent most likely had a greater predisposition to a complication developing, as the surgeons would typically place stents in more complex cases at risk for complications. Again, our study might have been underpowered to detect a difference in ureterovesical complications in the selective cohort only, as this constituted a smaller subset of our entire study sample. Despite the increased use of stents in the prophylactic group, we did not see an increase in either forgotten stents or UTIs after transplant. Our finding with respect to UTIs is contrary to what was reported in the meta-analysis by Wilson et al.12 In their meta-analysis, they identified a significant increase in UTIs associated with stents (risk ratio, 1.49; 95% CI, 1.04-2.15, P = .03). However, heterogeneity was significant (χ2 = 15.32, P = .02, I2 = 60.8%) in their pooled results for this outcome. Subsequent subgroup analysis revealed that the risk was associated with the antibiotic regime. Specifically, studies detailing prophylaxis with longterm cotrimoxazole had an equivalent risk ratio for
325
Ordon et al infection between patients with a stent and patients without a stent (risk ratio, 0.97; 95% CI, 0.71-1.33) compared with other regimes. Accordingly, at our institution, all transplant patients are maintained on longterm therapy with cotrimoxazole for a minimum of 1 year, and this fact might explain our findings. Nevertheless, the results of our secondary outcomes must be interpreted with caution, because our study was not powered to detect a difference in these outcomes. Our forgotten stent rate was higher than expected in both groups, highlighting the importance of a coordinated system to ensure that all patients return for stent removal. Despite some lengthy indwelling times for the forgotten stents, most patients (75%) did not require any additional intervention for stent removal beyond a cystoscopy. The absence of encrustation and stone formation due to the forgotten stent is in keeping with the previously reported low rates of allograft lithiasis.15-17 As previously mentioned, the use of prophylactic stents compared with no stent following kidney transplant has been well studied. In a meta-analysis published in 2004, Mangus and Haag11 reviewed 49 articles published between 1973 and 2002 evaluating the use of ureteric stents at the time of kidney transplant. The meta-analysis included 5 randomized controlled trials and 44 case series. In the controlled trials group, urologic complications occurred in 6 of 407 patients with a stent (1.5%) and 35 of 389 patients without a stent (9.0%; odds ratio, 0.24; 95% CI, 0.10-0.57; P
