Between May 23, 2010, and Nov 30, 2013, 227 participants were randomized into the study (n = 117, late group; n = 110, early group) (Figure 1). There were 15 postrandomization exclusions due to ineligibility, six were lost to follow-up (transferred care to another center), and one did not complete follow-up (died during study); in total, 205 participants were included in the final analysis for the primary outcome. In the early group, 21 patients failed to receive the allocated treatment and were regarded as crossovers into the late group, undergoing late stent removal. Baseline characteristics (Table 1) were similar between the allocation groups. Results below are reported according to treatment received (PP), with both PP and ITT results shown in Table 2. Of the 205 participants included in the final analysis, 169 completed the week 1 questionnaires and 158 completed the week 6 questionnaires; there was no difference in the proportion completed between groups at week 1 (p = 0.14) but a significant difference at week 6 (p = 0.04), with more completed in early removal group (data not shown). The median duration of the stent in the late group was 42 days (interquartile range [IQR] 38–47 days) compared with 5 days (IQR 5–5 days) in the early group.
Trial profile for primary outcome measures (per protocol analysis). ITT, intention to treat. [Color figure can be viewed at wileyonlinelibrary.com]
The primary outcome measure of stent complications (a composite of pain, UTI, hematuria, fragmentation, and migration) at 3 mo was significantly higher in the late versus early stent removal groups (36 of 126 vs. 6 of 79; p < 0.001) (Table 2). The adjusted analyses (adjusted for age, donor type, previous recurrent UTIs, dialysis status, diabetes status, and previous transplantation, in addition to treatment arm) found similar results (PP, p = 0.006; ITT, p = 0.099; results are available on request). The majority of stent complications consisted of UTIs, with incidence of 24% (31 of 126) in the late group compared with 7.6% (6 of 79) in the early group (p = 0.004). UTIs were mostly symptomatic (19 of 126, 15.1%) in the late group, with lower urinary tract symptoms more common than episodes of pyelonephritis (Table 3). In the early group, asymptomatic UTIs were more common (4 of 79, 5.1%) than symptomatic UTIs (2 of 79, 2.5%). The most common organisms were Escherichia coli (45%), Enterococcus faecalis (18%), and Klebsiella (11%). Other stent complications were seen only in the late group: pain in 4 of 126 (3.2%), hematuria in 2 of 126 (1.6%), and migration to the bladder in 3 of 126 (2.4%). There was no incidence of stent fragmentation or encrustation.
Secondary outcome measures included the incidence of MUCs (urine leak and/or stenosis), which was higher in the early group (3 of 80, 3.75%) compared with the late group (1 of 126, 0.8%) but was not statistically significant (p = 0.36). All four patients with MUCs required open surgical revision. There were three patients with transient obstruction in the early group.
Analysis of health status and QoL was assessed using FAIT-U and EQ-5D validated questionnaires. Although there was no difference between the groups at 1 week postoperatively, the FAIT-U total score as well as the physical and functional well-being subscales were significantly better in the early group at 6 weeks (p = 0.012) (Table 2 and Supplementary Data). No difference was seen between groups with EQ-5D questionnaires at either time frame. Diary cards were used as a subjective measure of patient acceptability (Supplementary Data), with no difference reported between the groups for the pain or anxiety of the stent-removal procedure.
SAEs were reported in 20 participants, 13 in the late group and 7 in the early group. There was one death during the study in the early group; this was attributed to a cardiac event. Graft loss was reported in eight patients, three in the late group and five in the early group. None of the SAEs were attributed to trial treatment.
Other outcome measures included stent symptoms, which occurred only in the late-removal group (11 of 126, 8.7%; p = 0.015). There appeared to be a difference in the timing of UTI episodes between groups, with a median of 43 days (IQR 16.50–58.25 days) in late group and 28 days (IQR 18–40 days) in the early group (Supplementary Data). There was no difference between the groups in length of operative time (p = 0.645), inpatient hospital stay (p = 0.086), rejection episodes (p = 0.858), or eGFR (mL/min per 1.73m2; p = 0.481) at 6 mo (Table 2).
A sensitivity analysis removing all crossovers from the analysis maintained a significant difference in primary outcome measures between groups, with stent complication in 28 of 107 (26.2%) in the late group compared with 6 of 79 (7.6%) in the early group (p = 0.002). In the subgroup analyses, there was no evidence that a prior history of recurrent UTIs (p = 0.63), surgeon experience (p = 0.79), or donor type (p = 0.59) influenced the primary outcome, after adjusting for treatment arm and age group. There was a trend in favor of early stent removal for the primary outcome in all subgroups (Figure 2). In the pediatric subgroup, stent complications occurred in 6 of 22 (27.3%) in the late group and 1 of 10 (10%) in the early group (p = 0.387); stent symptoms were reported only in the late group (1 of 22, 4.5%), and there were no MUCs.
American Journal of Transplantation. 2017;17(8):2129-2138. © 2017 Blackwell Publishing