Abstract
Rationale & Objective
Management of asymptomatic bacteriuria in kidney transplant recipients remains uncertain. Our main objective was to evaluate whether patient- or episode-related factors could help identify episodes of asymptomatic bacteriuria associated with a higher risk of being followed by a symptomatic urinary tract infection.
Study Design
A single-center, retrospective cohort study.
Settings & Participants
Consecutive patients who received a kidney transplantation between January 1, 2008, and April 26, 2016, and experienced ≥1 episode of asymptomatic bacteriuria during the first 3 years posttransplantation.
Exposure
Recipient age, sex, diabetes, donor type, retransplant, urological abnormalities, thymoglobulin for induction, time since transplant, presence of ureteral stent, prior acute rejection, leukocyturia, hematuria, urinary nitrites, bacterial strain, prior urinary tract infection, resistance to antibiotics, antibiotic treatment.
Outcomes
The primary outcome was the occurrence of symptomatic urinary tract infection. The secondary outcome was antibiotic treatment of asymptomatic bacteriuria.
Analytical Approach
A Cox regression survival analysis model accounting for recurrent events and a logistic regression model.
Results
From a cohort of 508 patients, we included 597 episodes of asymptomatic bacteriuria detected in 119 outpatients. Antibiotics were prescribed in 26% of these episodes. Overall, 56 (9%) of episodes were followed by a symptomatic urinary tract infection. Pretransplant diabetes (hazard ratio [HR], 4.28; 95% confidence intervals [CI], 2.40-7.61), leukocyturia or hematuria (HR, 2.24; 95% CI, 1.27-3.96), and the presence of a ureteral stent (HR, 3.40; 95% CI, 1.33-8.70) were associated with development of a clinical urinary tract infection in patients with asymptomatic bacteriuria.
Limitations
The small number of events limits complete multivariable adjustment. It also prevents us from drawing a definite conclusion about the importance of a number of independent variables as risk factors for the outcome.
Conclusions
The benefit of treating episodes of asymptomatic bacteriuria with high-risk characteristics should be investigated in future trials.
Index Words: Asymptomatic bacteriuria, kidney transplantation, urinary tract infection
Plain Language Summary
Urine infections are very frequent in patients who have received a kidney transplant. Bacteria are sometimes found in the urine of patients who report no symptoms. This is called asymptomatic bacteriuria. Asymptomatic bacteriuria is a risk factor for having an overt symptomatic urine infection, but previous small studies did not show any benefit of treatment. However, this may be different if only patients with asymptomatic bacteriuria at high risk of developing an overt infection were treated. In this study, we identified 3 features that put patients with asymptomatic bacteriuria at higher risk of developing overt infections: diabetes, presence of a stent in the ureter, and white or red blood cells in the urine when bacteriuria is detected. Examining the benefit of screening and treatment of asymptomatic bacteriuria in patients with these high-risk features warrants further investigation.
Introduction
While the recommended routine use of prophylactic antibiotics for 6-12 months after transplantation has helped in reducing the incidence of urinary tract infection (UTI) in kidney transplant recipients (KTRs),1, 2, 3, 4 the urinary tract remains the most common source of bacterial infection in these patients,5 with reported cumulative incidences of 30%-70% at 1-3 years follow-up.6, 7, 8, 9 Posttransplant UTIs are not only frequent but are also associated with adverse outcomes such as acute and chronic rejection, impaired long-term graft function, allograft loss, and death.10, 11, 12, 13, 14, 15, 16, 17 In the first month after kidney transplantation, bacteriuria is reported in more than half of screened patients, and up to 70% of these may be asymptomatic.9,18,19
Currently, there is insufficient evidence and no definitive recommendation to support the role of systematic screening and treatment of KTRs with asymptomatic bacteriuria.20 Recent clinical trials could not demonstrate that such a strategy provided clinical benefit.21,22 However, the study populations were not targeted high-risk groups, and the studies had limited power, with large confidence intervals (CIs) for the treatment effect that contained clinically important values. It is possible that treatment of asymptomatic bacteriuria may not be useful in KTRs at large but that specific patient- or episode-related characteristics may place patients at higher risk of developing clinical UTI after asymptomatic bacteriuria. Treatment of asymptomatic bacteriuria in such selected patients and episodes may diminish the number needed to treat to show benefit in terms of UTI prevention. We hypothesize that there are some characteristics, either patient- or episode-related, that are associated with an increased risk of asymptomatic bacteriuria progressing to symptomatic UTI.
In the present study, our objectives were (1) to identify patient- or episode-related characteristics that are associated with subsequent symptomatic UTI in KTRs who have asymptomatic bacteriuria; and (2) to understand the current prescription patterns by defining the factors associated with antibiotic treatment of asymptomatic bacteriuria in our KTR outpatient population.
Methods
Patients and Bacteriuric Episodes
We conducted a retrospective cohort study of patients who received a kidney transplant in a university-affiliated, Canadian adult transplant center between January 1, 2008, and April 26, 2016. Patients who received a kidney in combination with another solid organ transplant were excluded. Episodes of bacteriuria were collected for the first 3 years after transplantation. Bacteriuria was defined as a quantitative count of a bacterial strain exceeding 105 colony-forming units/mL in a urine culture. This is consistent with the definitions used by the Infectious Disease Society of America and European Association of Urology.23, 24, 25 Bacteriuric episodes occurring after graft failure were excluded, as well as asymptomatic bacteriuria or UTI supervening during a hospitalization for any other cause than a UTI. Cultures showing only fungi or normal genital/cutaneous flora (Corynebacterium spp., Staphylococcus epidermidis, Lactobacillus) were also excluded. Patient follow-up continued until 3 years posttransplantation, transfer from our center, the date of graft loss, or patient death, whichever occurred first. The project was approved by the hospital ethics committee (REB number 2015-5466). We followed the STROBE statement for reporting of cohort studies.26
Setting
Patients received triple immunosuppression typically consisting of tacrolimus, mycophenolate mofetil, and prednisone. Basiliximab was used as a standard induction therapy, whereas thymoglobulin was given to patients deemed at high immunological risk (calculated panel reactive antibody levels >80%, African American ethnicity, previous immune-mediated graft loss). At our center, corticosteroids are progressively tapered to 5 mg/d over 3 months and continued indefinitely. A ureteral stent is routinely inserted at the time of transplantation and removed 4-8 weeks later. Trimethoprim-sulfamethoxazole prophylaxis (80-400 mg twice weekly) is routinely prescribed for 1 year after transplantation to prevent Pneumocystis jirovecii infection. This prophylaxis was extended as long as the graft still functioned after 2013, when an increase in the incidence of late P jirovecii infections was observed.27 No other urinary prophylaxis was given. According to the local protocol, urine analyses and cultures are routinely conducted twice a week during the first month posttransplantation, once a week between months 2 and 3, once every 2 weeks between months 4 and 6, once a month between months 7-12, every second month during the second year posttransplantation, and at least quarterly from the third year posttransplantation onward. Further urinary analyses and cultures can be requested by transplant physicians if patients present with symptoms or following a previous asymptomatic bacteriuria episode. A positive culture, whether accompanied by symptoms or not, is always reported to the transplant physician. Antibiotic therapy is left to the discretion of this physician with no specific protocol in place to guide treatment.
Measurements
The primary outcome was the occurrence of a clinical UTI (cystitis or pyelonephritis) after an episode of asymptomatic bacteriuria. The diagnoses of cystitis (irritative urinary symptoms with a positive urine culture in the absence of fever) and graft pyelonephritis (fever and/or chills in the presence of a positive culture, graft tenderness, or irritative urinary symptoms and in the absence of an alternative diagnosis) were retrieved from patient charts and from a computerized clinical database, as documented by the treating physicians. The secondary outcome was an antibiotic prescription for the treatment of asymptomatic bacteriuria, retrieved from the patient’s charts.
Independent variables were considered as patient-related or episode-related. Patient-related variables were recipient age, sex, race, pretransplant diabetes, donor type (living vs deceased), retransplantation, the presence of a major urological abnormality (need for intermittent catheterization, postvoid residual urine levels >500 mL, surgical reconstruction of the urinary tract, or urethral strictures), and use of thymoglobulin as induction therapy. Episode-related characteristics included the time elapsed between transplantation and asymptomatic bacteriuria, the presence of a ureteral stent, biopsy-proven Banff grade ≥1 T-cell or antibody-mediated acute rejection in the preceding 3 months, leukocyturia or hematuria, urinary nitrates, bacterial strain (dichotomized as Gram-negative vs all others), previous posttransplant clinical UTI occurring during the study period, resistance to trimethoprim-sulfamethoxazole, ciprofloxacin, or extended spectrum β-lactamase (ESBL) production, and the prescription of an antibiotic for asymptomatic bacteriuria. There were no missing data.
Statistical Analyses
Continuous variables are reported as means and standard deviations or medians and interquartile ranges, depending on their distribution. Categorical variables are summarized as proportions. First, we examined the outcome of each asymptomatic bacteriuria episode. This outcome was divided into 5 classes based on the next urine culture: negative culture (clearance), persistent asymptomatic bacteriuria (with the same or with a different microorganism), or clinical UTI (with the same organism or with a different organism). We compared the relative frequencies of these 3 outcomes for asymptomatic bacteriuria episodes that were treated versus untreated with χ2 tests.
To determine the factors associated with the subsequent occurrence of a clinical UTI in patients with asymptomatic bacteriuria, we used a Cox regression survival analysis model accounting for recurrent events by using a sandwich estimator of the covariance matrix. The trajectory of asymptomatic bacteriuria episodes for the main analysis can be found in Fig 1. Patients entered the cohort on the date of the first asymptomatic bacteriuria episode. A subsequent positive urine culture at the next urine examination with symptoms of a UTI constituted an event, and follow-up time was censored at that point. When the next urine culture was negative or if asymptomatic bacteriuria was still detected, follow-up time for the first asymptomatic bacteriuria episode was censored without an event occurring. When the next culture was positive with symptoms (event) or negative (no event), patients could reenter the cohort on the date of a subsequent episode of asymptomatic bacteriuria, if the latter did occur during follow-up. When the next culture was positive in the absence of symptoms, the follow-up time for the initial episode of asymptomatic bacteriuria was censored with no event documented, and the patient reentered the cohort on the date of that culture, which was considered a new episode of asymptomatic bacteriuria.
Figure 1.
Trajectory of asymptomatic bacteriuria (AB) episodes for the main analysis. Patients enter the analytic cohort on the date of the first AB episode. Follow-up is censored on the date of the next urine culture. On this date, 3 scenarios can occur: 1) the next culture is positive with symptoms (symptomatic urinary tract infection, event); 2) the next culture is negative (clearance, no event); 3) the next culture shows AB with the same or a different microorganism (persistent AB, no event). In scenarios 1 and 2, patients can reenter the cohort on the date of a subsequent episode of AB. In scenario 3, the patient reenters the cohort on the date of the culture showing persistent AB and this constitutes a novel AB episode to be analyzed.
To determine the factors that were associated with an increased likelihood of receiving antibiotic treatment for an asymptomatic bacteriuria episode, we fitted a logistic regression model including the independent variables above. The clustering effect of repeated episodes in the same patient over time was accounted for using generalized estimating equations. For both models, we included in the multivariable analyses all the independent variables that were associated with the outcome in univariable analyses with a P value ≤ 0.10. To avoid overfitting, the independent variables included in the multivariable model were restricted to those showing at least a statistical trend for an association with the outcome, given the relatively small number of events. Hence, 6 variables were included in the initial multivariable model for the primary outcome. To avoid overfitting, we reduced this number to 5 by excluding 1 variable that was not statistically associated with the outcome and did not modify the regression coefficients for the other variables in the multivariable model. Statistical analyses were performed using SAS version 9.4.
Results
Among the 508 patients who received a kidney transplant during the study period, 248 never had documented bacteriuria on follow-up. After applying the various inclusion/exclusion criteria, the bacteriuria study cohort included 119 patients with a total of 597 episodes of asymptomatic bacteriuria (Fig 2).
Figure 2.
Flow chart of patients according to the inclusion/exclusion criteria of the study.
Patient-related characteristics of the population with asymptomatic bacteriuria are shown in Table 1. Episode-related characteristics are described in Table 2. Antibiotics were prescribed in 26% of asymptomatic bacteriuria episodes. The incidence of bacterial resistance to trimethoprim-sulfamethoxazole and/or ciprofloxacin and/or ESBL-producing strains was high at 64%. Gram-negative strains were isolated in 63% of cultures and Enterococci in 34%, with isolation of other microorganisms in only 4% of cultures. In 44% of asymptomatic bacteriuria episodes, there was at least 1 prior episode of a posttransplant symptomatic UTI. There were no missing data on the collected variables except for hematuria (n = 51) and leukocyturia (n = 19).
Table 1.
Characteristics of Patients Included in the Bacteriuria Study Cohort
| Characteristics | All Patients n = 119 | No Infection n = 84 | ≥1 Infection n = 35 |
|---|---|---|---|
| Age in years, mean (SD) | 52 (13) | 52 (13) | 53 (14) |
| Male sex, n (%) | 48 (40%) | 35 (42%) | 13 (37%) |
| Ethnic origin, n (%) | |||
| White | 99 (83%) | 71 (85%) | 28 (80%) |
| African American | 11 (9%) | 8 (10%) | 3 (9%) |
| Asian | 9 (8%) | 5 (6%) | 4 (11%) |
| Etiology of end-stage kidney disease, n (%) | |||
| Diabetes | 16 (13%) | 8 (10%) | 8 (23%) |
| Urological | 15 (13%) | 12 (14%) | 3 (9%) |
| Glomerular | 18 (15%) | 14 (17%) | 4 (11%) |
| Polycystic kidney disease | 14 (12%) | 12 (14%) | 2 (6%) |
| Other/unknown | 56 (47%) | 38 (45%) | 18 (51%) |
| Diabetes pretransplant, n (%) | 29 (24%) | 13 (15%) | 16 (46%) |
| First transplantations, n (%) | 102 (86%) | 73 (87%) | 29 (83%) |
| Living donor, n (%) | 22 (18%) | 18 (21%) | 4 (11%) |
| Induction with thymoglobulin, n (%) | 35 (29%) | 24 (29%) | 11 (31%) |
| Rejection episode(s), n (%) | 3 (3%) | 1 (1%) | 2 (6%) |
| Median duration of ureteral stent, d (IQR) | 42 (28-60) | 42 (26-59) | 45 (28-66) |
| Number of episodes of asymptomatic bacteriuria per patient, n (%) | |||
| 1-2 | 59 (50%) | 51 (61%) | 8 (23%) |
| 3-5 | 28 (23%) | 20 (24%) | 8 (23%) |
| >5 | 32 (27%) | 13 (15%) | 19 (54%) |
Abbreviations: IQR, interquartile range; SD, standard deviation.
Table 2.
Characteristics of Asymptomatic Bacteriuria Episodes
| Characteristics | All Episodes n = 597 | Episodes Without Infection n = 540 | Episodes With Infection n = 57 |
|---|---|---|---|
| Received antibiotic treatment, n (%) | 157 (26%) | 148 (27%) | 9 (16%) |
| Median time elapsed since transplant, d (IQR) | 393 (146-843) | 365 (139-810) | 656 (247-1049) |
| Presence of a ureteral stent, n (%) | 52 (9%) | 47 (9%) | 5 (9%) |
| Leukocyturia, n (%) | 352 (59%) | 310 (57%) | 42 (74%) |
| Hematuria, n (%) | 72 (12%) | 58 (12%) | 8 (15%) |
| Urinary nitrites, n (%) | 153 (26%) | 140 (26%) | 13 (23%) |
| Isolated microorganisms, n (%) | |||
| 1 Gram-negative strain | 331 (55%) | 296 (55%) | 35 (61%) |
| Escherichia coli | 210 | 191 | 19 |
| Klebsiella pneumoniae | 99 | 85 | 14 |
| Pseudomonas aeruginosa | 6 | 5 | 1 |
| Enterobacter cloacae | 4 | 3 | 1 |
| Citrobacter freundii | 3 | 3 | 0 |
| Citrobacter youngae | 3 | 3 | 0 |
| Klebsiella oxytoca | 2 | 2 | 0 |
| Proteus mirabilis | 2 | 2 | 0 |
| Citrobacter koseri | 1 | 1 | 0 |
| Pseudomonas putida | 1 | 1 | 0 |
| 1 Enterococcus strain | 194 (33%) | 179 (33%) | 15 (26%) |
| 1 strain of another bacterium | 21 (4%) | 19 (4%) | 2 (4%) |
| 2 Gram-negative strains | 43 (7%) | 38 (7%) | 5 (9%) |
| 2 Enterococcus strains | 6 (1%) | 6 (1%) | 0 (0%) |
| 2 strains of other bacteria | 2 (<1%) | 2 (<1%) | 0 (0%) |
| Presence of ≥1 resistant bacterium to cotrimoxazole or ciprofloxacin or extended spectrum β-lactamase strain, n (%) | 380 (64%) | 342 (63%) | 38 (67%) |
| Previous acute pyelonephritis or cystitis, n (%) | 264 (44%) | 225 (42%) | 39 (68%) |
| Rejection in the 3 months preceding the episode of asymptomatic bacteriuria, n (%) | 12 (2%) | 11 (2%) | 1 (2%) |
| Median time between initial asymptomatic bacteriuria and clinical urinary tract infection, d (IQR) | 14 (5-36) |
Table 3 shows the outcomes of asymptomatic bacteriuria in the treated and untreated episodes based on the clinical and microbiological profile of the next urine culture. Overall, 56 (9%) of episodes of asymptomatic bacteriuria were followed by a symptomatic UTI at a median of 14 days (interquartile range, 5-36 days) after the initial detection of asymptomatic bacteriuria, 45 (80%) of which were caused by the same microorganism. There was a significant difference in the outcome of asymptomatic bacteriuria based on treatment status. This difference was accounted for by more clearance of bacteriuria in the treated episodes (107/157, 68%) versus the untreated ones (96/440, 22%) (P < 0.01). However, there was no significant difference observed in the subsequent development of a symptomatic UTI between the treated and untreated groups (9/157 [6%] episodes for treated vs 47/440 [11%] for untreated, P = 0.09).
Table 3.
Outcomes of Asymptomatic Bacteriuria in the Treated and Untreated Episodes Based on the Clinical and Microbiological Profile of the Next Urine Culture
| Outcome of Asymptomatic Bacteriuria Episode (n = 597) | Treated (n = 157) | Untreated (n = 440) | Pa |
|---|---|---|---|
| Symptomatic urinary tract infection, n (%) | 9 (6%) | 47 (11%) | 0.09 |
| Same microorganism | 7 (4%) | 38 (9%) | |
| Different microorganism | 2 (1%) | 9 (2%) | |
| Persistent asymptomatic bacteriuria, n (%) | 41 (26%) | 297 (68%) | <0.01 |
| Same microorganism | 27 (17%) | 274 (62%) | |
| Different microorganism | 14 (9%) | 23 (5%) | |
| Negative urine culture, n (%) | 107 (68%) | 96 (22%) | <0.01 |
χ2 test.
In the multivariable model, the presence of leukocyturia or hematuria (hazard ratio [HR], 2.24; 95% CI, 1.27-3.96), pretransplant diabetes (HR, 4.28; 95% CI, 2.40-7.61), and the presence of a ureteral stent (HR, 3.40; 95% CI, 1.33-8.70) were indicative of an increased likelihood of developing a subsequent symptomatic UTI after an episode of asymptomatic bacteriuria. Prior clinical UTI was associated with an increased risk of developing a subsequent symptomatic UTI after an episode of asymptomatic bacteriuria in the univariable analysis (HR, 2.38; 95% CI, 1.19-4.76), but this was no longer statistically significant in the multivariable model (HR, 1.70; 95% CI, 0.95-3.05) (Table 4). Antibiotic therapy for asymptomatic bacteriuria was not associated with a decreased rate of subsequent UTI in univariable analyses.
Table 4.
Factors Associated With the Development of a Symptomatic Urinary Tract Infection Following Asymptomatic Bacteriuria
| Variable | Univariable Analyses |
Multivariable Analyses |
||
|---|---|---|---|---|
| HR (95% CI) | P | HR (95% CI) | P | |
| Pretransplant diabetes | 4.81 (2.63, 8.78) | <0.01 | 4.28 (2.40, 7.61) | <0.01 |
| Leukocyturia or hematuria | 2.31 (1.23, 4.35) | <0.01 | 2.24 (1.27, 3.96) | <0.01 |
| Previous acute pyelonephritis or cystitis | 2.38 (1.19, 4.76) | 0.01 | 1.70 (0.95, 3.05) | 0.08 |
| Presence of a ureteral stent | 4.16 (1.60, 10.85) | <0.01 | 3.40 (1.33, 8.70) | 0.01 |
| Previous transplantations | 2.86 (1.28, 6.45) | 0.01 | 1.76 (0.88, 3.54) | 0.11 |
| Female sex | 2.06 (0.95, 4.47) | 0.07 | - | - |
| Age (per year) | 0.98 (0.95, 1.01) | 0.13 | - | - |
| Ethnic origin | - | - | ||
| White (reference) | 1.00 | |||
| African American | 1.25 (0.32, 4.89) | 0.75 | ||
| Asian | 2.13 (0.83, 5.43) | 0.11 | ||
| Living donor | 0.94 (0.38, 2.32) | 0.54 | - | - |
| Thymoglobulin at induction | 1.46 (0.71, 3.02) | 0.31 | - | - |
| Time elapsed since transplant (per year) | 1.02 (0.79, 1.31) | 0.88 | - | - |
| Rejection episode within 3 mo before asymptomatic bacteriuria episode | 1.84 (0.31, 10.84) | 0.50 | - | - |
| Urinary nitrates | 0.67 (0.31, 1.48) | 0.32 | - | - |
| Gram-negative strain | 1.11 (0.47, 2.61) | 0.81 | ||
| Resistance to cotrimoxazole, ciprofloxacin or ESBL-producing strain | 0.90 (0.48, 1.69) | 0.75 | - | - |
| Number of previous episodes of asymptomatic bacteriuria | - | - | ||
| 0-1 | 1.00 | |||
| 2-4 | 0.75 (0.29, 1.96) | 0.56 | ||
| ≥5 | 0.64 (0.29, 1.43) | 0.28 | ||
| Major urological abnormality | 0.59 (0.25, 1.36) | 0.21 | - | - |
| Antibiotic treatment of the asymptomatic bacteriuria episode | 0.58 (0.24-1.38) | 0.21 | - | - |
Abbreviations: CI, confidence interval; ESBL, extended spectrum β-lactamase; HR, hazard ratio.
Factors found to be associated with prescribing antibiotic treatment for an asymptomatic bacteriuria episode are shown in Table 5. In multivariable analyses, the presence of leukocyturia or hematuria was associated with increased likelihood of antibiotic prescription (odds ratio [OR], 1.96; 95% CI, 1.26-3.05), as were positive cultures showing a Gram-negative strain (OR, 4.21; 95% CI, 2.47-7.20). Resistance to cotrimoxazole, ciprofloxacin, or ESBL-producing strain (OR, 0.47; 95% CI, 0.27-0.82) and an elevated number of asymptomatic bacteriuria episodes since transplantation (for ≥5 vs 0-1 previous episodes: OR, 0.41; 95% CI, 0.22-0.77) were, for their part, associated with a decreased likelihood of prescribing antibiotic treatment. Although they were associated with an increased likelihood of developing symptomatic UTI after asymptomatic bacteriuria, pretransplant diabetes and the presence of a ureteral stent were not associated with antibiotic prescription. In contrast, the presence of leukocyturia or hematuria was associated with both an increased likelihood of antibiotic treatment and the development of symptomatic UTI.
Table 5.
Factors Associated With Receiving Antibiotic Treatment for Asymptomatic Bacteriuria
| Variables | Univariable analyses |
Multivariable analyses |
||
|---|---|---|---|---|
| OR (95% CI) | P | OR (95% CI) | P | |
| Leukocyturia or hematuria | 1.70 (1.08, 2.68) | 0.03 | 1.96 (1.26, 3.05) | <0.01 |
| Gram-negative strain | 3.48 (1.86, 6.50) | <0.01 | 4.21 (2.47, 7.20) | <0.01 |
| Resistance to cotrimoxazole, ciprofloxacin, or ESBL-producing strain | 0.62 (0.36, 1.08) | 0.09 | 0.47 (0.27, 0.82) | <0.01 |
| Number of previous episodes of asymptomatic bacteriuria | ||||
| 0-1 | 1.00 | 1.00 | ||
| 2-4 | 1.21 (0.71, 2.07) | 0.48 | 1.20 (0.66, 2.19) | 0.55 |
| ≥5 | 0.38 (0.21, 0.71) | <0.01 | 0.41 (0.22, 0.77) | <0.01 |
| Female sex | 2.49 (1.33, 4.68) | <0.01 | 1.53 (0.88, 2.65) | 0.13 |
| Age (per year) | 0.97 (0.95, 0.99) | <0.01 | 0.99 (0.98, 1.00) | 0.18 |
| Major urological pathology | 0.26 (0.08, 0.84) | 0.02 | 0.34 (0.11, 1.06) | 0.06 |
| Ethnic origin | ||||
| White | 1.00 | 1.00 | ||
| African American | 1.11 (0.52, 2.37) | 0.78 | 1.23 (0.57, 2.67) | 0.60 |
| Asian | 1.97 (1.08, 3.58) | 0.03 | 1.23 (0.66, 2.30) | 0.52 |
| Pretransplant diabetes | 1.08 (0.62, 1.88) | 0.80 | - | - |
| Previous acute pyelonephritis or cystitis | 1.50 (0.78, 2.86) | 0.22 | - | - |
| Presence of a ureteral stent | 1.04 (0.56, 1.93) | 0.91 | - | - |
| Previous transplantations | 1.71 (0.84, 3.48) | 0.14 | - | - |
| Living donor | 1.50 (0.75, 3.02) | 0.25 | - | - |
| Thymoglobulin at induction | 1.68 (0.88, 3.21) | 0.12 | - | - |
| Time elapsed since transplant (per year) | 0.95 (0.81, 1.11) | 0.52 | - | - |
| Rejection episode within 3 mo before asymptomatic bacteriuria episode | 0.93 (0.20, 4.32) | 0.93 | - | - |
| Urinary nitrates | 1.40 (0.68, 2.88) | 0.36 | - | - |
Abbreviations: CI, confidence interval; ESBL, extended spectrum β-lactamase; OR, odds ratio.
Discussion
Current recommendations and recent randomized controlled trials do not support routine treatment of asymptomatic bacteriuria in KTRs in general.20, 21, 22 However, certain subsets of patients or episodes may have features that put them at high risk of developing a clinical UTI. In these “high-risk” subsets, the benefit of antibiotic treatment may be substantially different and would warrant further investigation. In this study, we sought to define such high-risk features and to understand the factors that currently influence antibiotic therapy at our center. We found that pretransplant diabetes, the presence of leukocyturia or hematuria, and the presence of a ureteral stent were associated with an increased likelihood of developing a clinical UTI after asymptomatic bacteriuria. We also observed that prescribing antibiotic treatment for asymptomatic bacteriuria was more likely in the presence of leukocyturia or hematuria and when a Gram-negative strain was isolated, whereas a decreased likelihood of receiving treatment was observed when antibiotic resistance was present, as well as when patients had multiple asymptomatic bacteriuria episodes in the past.
In KTRs, symptomatic UTIs constitute a significant burden in terms of morbidity, drug therapy, and hospitalization costs.11,28 Although UTIs in KTRs have not been directly associated with increased mortality from urinary sepsis, blood stream infections, particularly Gram-negative infections most frequently caused by a UTI, have been associated with all-cause mortality and with death-censored graft failure.14,29 Acute pyelonephritis was shown to be associated with acute rejection episodes in some studies,10,30 and a higher incidence of UTIs in KTRs has likewise been correlated with biopsy-proven chronic rejection.16 Furthermore, the recognition of subclinical UTIs as seen in pathologically diagnosed allograft pyelonephritis suggests that the extent and effects of allograft infection may be underrated.31 Acute graft pyelonephritis, particularly within the first 3 months posttransplantation, was also found to be associated with poorer graft survival in some studies32,33 but not in others.10,34, 35, 36
Asymptomatic bacteriuria has clearly been shown to be a risk factor for UTIs in KTRs, but whether treating asymptomatic bacteriuria prevents UTI is unclear.11,12,37, 38, 39 For instance, Fiorante et al detected asymptomatic bacteriuria in half of KTRs during the first 3 years posttransplantation, with a 7-fold increase in the incidence of pyelonephritis in patients with asymptomatic bacteriuria than in those without it, despite routine antibiotic treatment.11 Another study also found that asymptomatic bacteriuria was an independent risk factor for symptomatic UTIs despite universal antibiotic therapy. However, only 14% of symptomatic UTIs were preceded by asymptomatic bacteriuria with the same microorganism.37 Previous studies have identified multiple, often overlapping, risk factors for asymptomatic bacteriuria and/or UTI in KTRs such as ureteral stent insertion, induction therapy, female sex, advanced age, diabetes mellitus, underlying urinary tract abnormalities, graft rejection, and cadaveric donor kidneys as opposed to living donor kidney transplants.10,11,37,40 However, to our knowledge, this is the first study to identify features that predict a higher risk of developing clinical UTI after asymptomatic bacteriuria.
Two previous randomized controlled trials found no apparent benefit of treatment in KTRs.21,22 However, the reported CIs for the effect of treatment were wide, with lower boundaries of 0.11 and 0.50, which would represent clear clinical benefit. Hence, although the studies were negative, they did not exclude benefit given their low power. In another quasi-randomized prospective study, clinical UTI developed in 21% of treated KTRs with asymptomatic bacteriuria and pyuria after the first posttransplant year compared to 31% of untreated patients.38 Such a risk reduction would represent a clinically important OR of 0.68 in favor of treatment. Although CIs were not reported, this study again lacks power to draw a conclusion on the benefit of treatment. In a retrospective study that found no benefit of treating asymptomatic bacteriuria due to Escherichia coli or Enterococci from the first to the sixth month posttransplantation, a tendency to more frequent microbiological clearance with antibiotic therapy in those with pyuria and high-grade bacteriuria was reported.12 Another study even reported a higher incidence of UTI in patients with asymptomatic bacteriuria who received antibiotic treatment.39 Since the latter study did not attempt to determine the characteristics associated with receiving antibiotic treatment, their results could be due to confounding by indication, a frequent limitation in evaluating the effect of treatment in observational studies.
Along the same lines, we found that treatment of asymptomatic bacteriuria was associated with a greater likelihood of bacterial clearance on the next urine culture but was not protective against the development of subsequent UTI when all episodes were considered. When left untreated, 11% of episodes of asymptomatic bacteriuria converted to a clinical UTI on follow-up. Although this argues in favor of not treating all asymptomatic episodes, treatment in the presence of high-risk features for subsequent UTI such as leukocyturia/hematuria, presence of a ureteral stent, and diabetes could yield different results and warrants further investigation. Selecting high-risk asymptomatic bacteriuria episodes for inclusion in future trials would increase the number of events and improve the power to detect a difference in outcome by antibiotic treatment.
The management of asymptomatic bacteriuria in KTRs presents an ongoing dilemma. On the one hand, treatment of asymptomatic bacteriuria raises concerns about increasing antibiotic resistance or medication side effects and/or interactions. On the other hand, the transplant setting is associated with structural aberrancy and immunosuppression, in which asymptomatic and symptomatic bacteriuria can be associated with renal scarring due to a degree of vesicoureteral reflux.41 Adding to the controversy is the potential of antibiotic therapy to compromise the microbial balance conferred by less virulent colonizing bacteria that prevent superinfection with more pathogenic strains when left untreated.42 During the study period, we performed 19,304 per protocol urine cultures in our 508 study participants and identified 597 asymptomatic bacteriuria episodes occurring in 119 patients. The cost-effectiveness of this screening procedure is questionable. The optimal time window for asymptomatic bacteriuria screening is not clear, but several authors suggest screening for and treating asymptomatic bacteriuria in the first 3 months after kidney transplantation. Consensus is lacking on the role of subsequent screening and management of asymptomatic bacteriuria in this population.43,44 A targeted approach to screening for asymptomatic bacteriuria in those at higher risk of developing a clinical UTI (presence of a ureteral stent, diabetes, or hematuria/leukocyturia on the urine analysis) warrants further study.
Due to lack of specific standardized protocols, management practices vary even within institutions. For instance, at our center, management is at the discretion of transplant physicians. Hence, we sought to determine the practice patterns for the treatment of asymptomatic bacteriuria. We found that although patients were more likely to be treated for asymptomatic bacteriuria in the presence of leukocyturia or hematuria, which is associated with a higher risk of developing subsequent UTI, they were not more likely to be treated in diabetic patients or those with a ureteral stent, factors that were also associated with an increased likelihood of subsequent UTI. Concerns regarding the development of antibiotic resistance may explain the lower likelihood of treatment of asymptomatic bacteriuria with multiple previous episodes or with strains that already have a resistance profile.
Our study is limited by the small number of events (symptomatic UTIs), which limits multivariable adjustment. It also prevents us from drawing a definite conclusion as to the importance of a number of independent variables as risk factors for the outcome. For example, allograft rejection, induction therapy, and female sex had effect sizes that would have been clinically important, but CIs were wide due to limited power. It is also a single-center study including a predominantly White, first transplant recipient cohort, which limits the generalizability of our findings. Last, we did not examine the associations between the treatment of asymptomatic bacteriuria and allograft rejection, failure, or death. Hence, we cannot draw any conclusion as to whether treating asymptomatic bacteriuria could have prevented these outcomes.
In summary, the current literature indicates that KTRs with asymptomatic bacteriuria are more likely to develop subsequent symptomatic UTI regardless of whether or not antibiotics were prescribed for the subclinical infection. However, the limited power of previous studies precludes a definite conclusion on the efficacy of antimicrobial treatment in this patient population. We have identified pretransplant diabetes, leukocyturia/hematuria, and the presence of a ureteral stent as predictors of an increased likelihood of a symptomatic UTI after asymptomatic bacteriuria. Some, but not all, of these features were also associated with an increased likelihood of receiving an antibiotic prescription. Distinguishing which episodes of asymptomatic bacteriuria have an increased risk for the development of symptomatic UTI based on easily identifiable and measurable parameters is a reasonable approach that can help in elucidating treatment effect, or lack thereof. We believe that the role of screening and treatment for asymptomatic bacteriuria is worth investigating through randomized studies in subsets of KTRs with high-risk features to answer this important clinical question.
Article Information
Authors’ Full Names and Academic Degrees
Samar Medani, MD, Marc Dorais, MSc, Aurélie Poulin, MD, Alexandre Tavares-Brum, BSc, Habib Mawad, MD, MSc, Alain Duclos, MD, Azemi Barama, MD, and Héloïse Cardinal, MD, PhD
Contributions
Research idea and study design: SM, HC, AD, AB; Data acquisition: SM, AT-B, HC, AD, AP, HM; Data analysis/interpretation: SM, HC, AB, AD, AP, HM; Statistical analysis: MD; Supervision or mentorship: HC. Each author contributed important intellectual content during article drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved.
Support
This work was supported by a donation from a CHUM patient, Ms. Chantale Caron. HC co-holds the Shire Chair in Nephrology, Transplantation and Renal Regeneration.
Financial Disclosure
The authors declare that they have no competing interests.
Acknowledgements
The investigators would like to acknowledge the CHUM kidney transplant nursing team for their work and support of the research activities.
Peer Review
Received February 20, 2024. Evaluated by 2 external peer reviewers, with direct editorial input from the Statistical Editor, an Associate Editor, and the Editor-in-Chief. Accepted in revised form September 27, 2024.
Footnotes
Complete author and article information provided before references.
References
- 1.Tolkoff-Rubin N.E., Cosimi A.B., Russell P.S., Rubin R.H. A controlled study of trimethoprim-sulfamethoxazole prophylaxis of urinary tract infection in renal transplant recipients. Rev Infect Dis. 1982;4(2):614–618. doi: 10.1093/clinids/4.2.614. [DOI] [PubMed] [Google Scholar]
- 2.Fox B.C., Sollinger H.W., Belzer F.O., Maki D.G. A prospective, randomized, double-blind study of trimethoprim-sulfamethoxazole for prophylaxis of infection in renal transplantation: clinical efficacy, absorption of trimethoprim-sulfamethoxazole, effects on the microflora, and the cost-benefit of prophylaxis. Am J Med. 1990;89(3):255–274. doi: 10.1016/0002-9343(90)90337-d. [DOI] [PubMed] [Google Scholar]
- 3.Green H., Rahamimov R., Gafter U., Leibovitci L., Paul M. Antibiotic prophylaxis for urinary tract infections in renal transplant recipients: a systematic review and meta-analysis. Transpl Infect Dis. 2011;13(5):441–447. doi: 10.1111/j.1399-3062.2011.00644.x. [DOI] [PubMed] [Google Scholar]
- 4.Kasiske B.L., Zeier M.G., Chapman J.R., et al. KDIGO clinical practice guideline for the care of kidney transplant recipients: a summary. Kidney Int. 2010;77(4):299–311. doi: 10.1038/ki.2009.377. [DOI] [PubMed] [Google Scholar]
- 5.Alangaden G.J., Thyagarajan R., Gruber S.A., et al. Infectious complications after kidney transplantation: current epidemiology and associated risk factors. Clin Transplant. 2006;20(4):401–409. doi: 10.1111/j.1399-0012.2006.00519.x. [DOI] [PubMed] [Google Scholar]
- 6.Sorto R., Irizar S.S., Delgadillo G., Alberú J., Correa-Rotter R., Morales-Buenrostro L.E. Risk factors for urinary tract infections during the first year after kidney transplantation. Transplant Proc. 2010;42(1):280–281. doi: 10.1016/j.transproceed.2009.11.029. [DOI] [PubMed] [Google Scholar]
- 7.Galindo Sacristán P., Pérez Marfil A., Osorio Moratalla J.M., et al. Predictive factors of infection in the first year after kidney transplantation. Transplant Proc. 2013;45(10):3620–3623. doi: 10.1016/j.transproceed.2013.11.009. [DOI] [PubMed] [Google Scholar]
- 8.Gołębiewska J.E., Dębska-Ślizień A., Rutkowski B. Urinary tract infections during the first year after renal transplantation: one center’s experience and a review of the literature. Clin Transplant. 2014;28(11):1263–1270. doi: 10.1111/ctr.12465. [DOI] [PubMed] [Google Scholar]
- 9.Maraha B., Bonten H., van Hooff H., Fiolet H., Buiting A.G., Stobberingh E.E. Infectious complications and antibiotic use in renal transplant recipients during a 1-year follow-up. Clin Microbiol Infect. 2001;7(11):619–625. doi: 10.1046/j.1198-743x.2001.00329.x. [DOI] [PubMed] [Google Scholar]
- 10.Pellé G., Vimont S., Levy P.P., et al. Acute pyelonephritis represents a risk factor impairing long-term kidney graft function. Am J Transplant. 2007;7(4):899–907. doi: 10.1111/j.1600-6143.2006.01700.x. [DOI] [PubMed] [Google Scholar]
- 11.Fiorante S., López-Medrano F., Lizasoain M., et al. Systematic screening and treatment of asymptomatic bacteriuria in renal transplant recipients. Kidney Int. 2010;78(8):774–781. doi: 10.1038/ki.2010.286. [DOI] [PubMed] [Google Scholar]
- 12.El Amari E.B., Hadaya K., Bühler L., et al. Outcome of treated and untreated asymptomatic bacteriuria in renal transplant recipients. Nephrol Dial Transplant. 2011;26(12):4109–4114. doi: 10.1093/ndt/gfr198. [DOI] [PubMed] [Google Scholar]
- 13.Abbott K.C., Swanson S.J., Richter E.R., et al. Late urinary tract infection after renal transplantation in the United States. Am J Kidney Dis. 2004;44(2):353–362. doi: 10.1053/j.ajkd.2004.04.040. [DOI] [PubMed] [Google Scholar]
- 14.Al-Hasan M.N., Razonable R.R., Kremers W.K., Baddour L.M. Impact of Gram-negative bloodstream infection on long-term allograft survival after kidney transplantation. Transplantation. 2011;91(11):1206–1210. doi: 10.1097/TP.0b013e3182180535. [DOI] [PubMed] [Google Scholar]
- 15.Lee J.R., Bang H., Dadhania D., et al. Independent risk factors for urinary tract infection and for subsequent bacteremia or acute cellular rejection: a single-center report of 1166 kidney allograft recipients. Transplantation. 2013;96(8):732–738. doi: 10.1097/TP.0b013e3182a04997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Müller V., Becker G., Delfs M., Albrecht K.H., Philipp T., Heemann U. Do urinary tract infections trigger chronic kidney transplant rejection in man? J Urol. 1998;159(6):1826–1829. doi: 10.1016/S0022-5347(01)63165-3. [DOI] [PubMed] [Google Scholar]
- 17.Ito K., Goto N., Futamura K., et al. Death and kidney allograft dysfunction after bacteremia. Clin Exp Nephrol. 2016;20(2):309–315. doi: 10.1007/s10157-015-1155-6. [DOI] [PubMed] [Google Scholar]
- 18.Gołębiewska J., Dębska-Ślizień A., Komarnicka J., Samet A., Rutkowski B. Urinary tract infections in renal transplant recipients. Transplant Proc. 2011;43(8):2985–2990. doi: 10.1016/j.transproceed.2011.07.010. [DOI] [PubMed] [Google Scholar]
- 19.Singh R., Geerlings S.E., Bemelman F.J. Asymptomatic bacteriuria and urinary tract infections among renal allograft recipients. Curr Opin Infect Dis. 2015;28(1):112–116. doi: 10.1097/QCO.0000000000000120. [DOI] [PubMed] [Google Scholar]
- 20.Nicolle L.E., Bradley S., Colgan R., et al. Infectious Diseases Society of America guidelines for the diagnosis and treatment of asymptomatic bacteriuria in adults. Clin Infect Dis. 2005;40(5):643–654. doi: 10.1086/427507. [DOI] [PubMed] [Google Scholar]
- 21.Origüen J., López-Medrano F., Fernández-Ruiz M., et al. Should asymptomatic bacteriuria be systematically treated in kidney transplant recipients? Results from a randomized controlled trial. Am J Transplant. 2016;16(10):2943–2953. doi: 10.1111/ajt.13829. [DOI] [PubMed] [Google Scholar]
- 22.Coussement J., Kamar N., Matignon M., et al. Antibiotics versus no therapy in kidney transplant recipients with asymptomatic bacteriuria (BiRT): a pragmatic, multicentre, randomized, controlled trial. Clin Microbiol Infect. 2021;27(3):398–405. doi: 10.1016/j.cmi.2020.09.005. [DOI] [PubMed] [Google Scholar]
- 23.Hooton T.M., Stamm W.E. Diagnosis and treatment of uncomplicated urinary tract infection. Infect Dis Clin North Am. 1997;11(3):551–581. doi: 10.1016/s0891-5520(05)70373-1. [DOI] [PubMed] [Google Scholar]
- 24.Warren J.W., Abrutyn E., Hebel J.R., Johnson J.R., Schaeffer A.J., Stamm W.E. Guidelines for antimicrobial treatment of uncomplicated acute bacterial cystitis and acute pyelonephritis in women. Infectious Diseases Society of America (IDSA) Clin Infect Dis. 1999;29(4):745–758. doi: 10.1086/520427. [DOI] [PubMed] [Google Scholar]
- 25.Grabe M., Bartoletti R., Bjerklund Johansen T.E., et al. Guidelines on urological infections. http://uroweb.org/wp-content/uploads/19-Urological-infections_LR2.pdf
- 26.von Elm E., Altman D.G., Egger M., et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med. 2007;147(8):573–577. doi: 10.7326/0003-4819-147-8-200710160-00010. [DOI] [PubMed] [Google Scholar]
- 27.Hosseini-Moghaddam S.M., Shokoohi M., Singh G., et al. A multicenter case-control study of the effect of acute rejection and cytomegalovirus infection on pneumocystis pneumonia in solid organ transplant recipients. Clin Infect Dis. 2019;68(8):1320–1326. doi: 10.1093/cid/ciy682. [DOI] [PubMed] [Google Scholar]
- 28.Becerra M.B., Becerra B.J., Safdar N. A nationwide assessment of asthma-mental health nexus among veterans. J Asthma. 2016;53(2):164–169. doi: 10.3109/02770903.2015.1086785. [DOI] [PubMed] [Google Scholar]
- 29.Silva M Jr, Marra A.R., Pereira C.A.P., Medina-Pestana J.O., Camargo L.F.A. Bloodstream infection after kidney transplantation: epidemiology, microbiology, associated risk factors, and outcome. Transplantation. 2010;90(5):581–587. doi: 10.1097/TP.0b013e3181e8a680. [DOI] [PubMed] [Google Scholar]
- 30.Kamath N.S., John G.T., Neelakantan N., Kirubakaran M.G., Jacob C.K. Acute graft pyelonephritis following renal transplantation. Transpl Infect Dis. 2006;8(3):140–147. doi: 10.1111/j.1399-3062.2006.00148.x. [DOI] [PubMed] [Google Scholar]
- 31.Varma P.P., Jairam A., Hooda A.K., Badwal S. Renal outcome in biopsy proven cases of graft pyelonephritis. Indian J Nephrol. 2014;24(3):161–165. doi: 10.4103/0971-4065.132009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Bodro M., Sanclemente G., Lipperheide I., et al. Impact of urinary tract infections on short-term kidney graft outcome. Clin Microbiol Infect. 2015;21(12):1104.e1–1104.e8. doi: 10.1016/j.cmi.2015.07.019. [DOI] [PubMed] [Google Scholar]
- 33.Giral M., Pascuariello G., Karam G., et al. Acute graft pyelonephritis and long-term kidney allograft outcome. Kidney Int. 2002;61(5):1880–1886. doi: 10.1046/j.1523-1755.2002.00323.x. [DOI] [PubMed] [Google Scholar]
- 34.Papasotiriou M., Savvidaki E., Kalliakmani P., et al. Predisposing factors to the development of urinary tract infections in renal transplant recipients and the impact on the long-term graft function. Ren Fail. 2011;33(4):405–410. doi: 10.3109/0886022X.2011.568137. [DOI] [PubMed] [Google Scholar]
- 35.Gołębiewska J., Dębska-Ślizień A., Zadrożny D., Rutkowski B. Acute graft pyelonephritis during the first year after renal transplantation. Transplant Proc. 2014;46(8):2743–2747. doi: 10.1016/j.transproceed.2014.09.076. [DOI] [PubMed] [Google Scholar]
- 36.Fiorante S., Fernández-Ruiz M., López-Medrano F., et al. Acute graft pyelonephritis in renal transplant recipients: incidence, risk factors and long-term outcome. Nephrol Dial Transplant. 2011;26(3):1065–1073. doi: 10.1093/ndt/gfq531. [DOI] [PubMed] [Google Scholar]
- 37.Gołębiewska J.E., Dębska-Ślizień A., Rutkowski B. Treated asymptomatic bacteriuria during first year after renal transplantation. Transpl Infect Dis. 2014;16(4):605–615. doi: 10.1111/tid.12255. [DOI] [PubMed] [Google Scholar]
- 38.Moradi M., Abbasi M., Moradi A., Boskabadi A., Jalali A. Effect of antibiotic therapy on asymptomatic bacteriuria in kidney transplant recipients. Urol J. 2005;2(1):32–35. [PubMed] [Google Scholar]
- 39.Green H., Rahamimov R., Goldberg E., et al. Consequences of treated versus untreated asymptomatic bacteriuria in the first year following kidney transplantation: retrospective observational study. Eur J Clin Microbiol Infect Dis. 2013;32(1):127–131. doi: 10.1007/s10096-012-1727-2. [DOI] [PubMed] [Google Scholar]
- 40.Lorenz E.C., Cosio F.G. The impact of urinary tract infections in renal transplant recipients. Kidney Int. 2010;78(8):719–721. doi: 10.1038/ki.2010.219. [DOI] [PubMed] [Google Scholar]
- 41.Ciszek M., Paczek L., Bartłomiejczyk I., Mucha K. Urine cytokines profile in renal transplant patients with asymptomatic bacteriuria. Transplantation. 2006;81(12):1653–1657. doi: 10.1097/01.tp.0000226072.20185.f8. [DOI] [PubMed] [Google Scholar]
- 42.Coussement J., Abramowicz D. Should we treat asymptomatic bacteriuria after renal transplantation? Nephrol Dial Transplant. 2014;29(2):260–262. doi: 10.1093/ndt/gft432. [DOI] [PubMed] [Google Scholar]
- 43.Yacoub R., Akl N.K. Urinary tract infections and asymptomatic bacteriuria in renal transplant recipients. J Glob Infect Dis. 2011;3(4):383–389. doi: 10.4103/0974-777X.91064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Parasuraman R., Julian K., AST Infectious Diseases Community of Practice Urinary tract infections in solid organ transplantation. Am J Transplant. 2013;13(suppl 4):327–336. doi: 10.1111/ajt.12124. [DOI] [PubMed] [Google Scholar]