Long-Term Kidney Outcomes in Children with Posterior Urethral Valves

Visual Abstract

Abstract

Key Points

• Among 727 children with posterior urethral valves, 32% had major adverse kidney events (death, kidney failure, or CKD) over a median of 14.2-year follow-up.

• Children with posterior urethral valves were at substantially higher risks of kidney failure, CKD, and hypertension than the general population.

• This justifies close kidney health surveillance among children with posterior urethral valves and optimized transitions to adult urologic care.

Background

Posterior urethral valves represent the most common cause of lower urinary tract obstruction in male infants (approximately 1/4000 live births). Long-term kidney outcomes of posterior urethral valves remain uncertain. We aimed to determine the time-varying risk of major adverse kidney events (MAKE) following children with posterior urethral valves into adulthood.

Methods

A population-based retrospective cohort study of all male children (<2 years) diagnosed with posterior urethral valves between 1991 and 2021 in Ontario, Canada. Comparator cohorts were (1) male general population and (2) male children with pyeloplasty (both <2 years). The primary outcome was MAKE (death, long-term KRT [dialysis or kidney transplant], or CKD). Time to MAKE was analyzed using multivariable-adjusted Cox proportional hazards models. We censored for provincial emigration or administrative censoring (March 31, 2022).

Results

We included 727 children with posterior urethral valves, 855 pyeloplasty comparators, and 1,013,052 general population comparators. The median follow-up time was 16.6 years (Q1–3, 8.6–24.5) overall. Throughout follow-up, 32% of children with posterior urethral valves developed MAKE versus 1% of the general population and 6% of pyeloplasty comparators. Their adjusted hazard ratio for MAKE was 36.6 (95% confidence interval, 31.6 to 42.4) versus the general population. The risk of developing MAKE declined over the first 5 years after posterior urethral valve diagnosis but remained elevated for >30-year follow-up. Children with posterior urethral valves were also at higher risk of death, CKD, long-term KRT, hypertension, and AKI than the general population or pyeloplasty comparators.

Conclusions

Children with posterior urethral valves are at higher risk of adverse long-term kidney outcomes well into adulthood.

Introduction

Posterior urethral valves are the most common cause of lower urinary tract obstruction in male children and are associated with significant morbidity, life-threatening complications, and lifelong disability.¹ Posterior urethral valves occur due to disrupted embryological urethral development, resulting in membranous folds obstructing the posterior urethral lumen, which create a high-pressure bladder and upper urinary tracts. Deranged kidney development from a common embryological insult or pressure-related injury causes irreversible kidney damage, dysplasia, tubular atrophy, and interstitial fibrosis. Bladder hypertrophy and fibrotic remodeling result in a small, hypercontractile, dysfunctional bladder during infancy that gradually evolves into a large-capacity, hypocontractile bladder in adolescence, both associated with poor kidney outcomes.^2,3 Posterior urethral valves are reported in 1 in 3800–8000 live births,^4–6 although the incidence is likely underestimated because half of prenatally detected fetuses undergo pregnancy termination or experience fetal demise.^7,8

Despite the use of prenatal and postnatal interventions to decompress the urinary tract, the long-term kidney outcomes of posterior urethral valves are poor. Observational studies report that 32%–54% of children with posterior urethral valves develop CKD and 15%–23% develop kidney failure.^9–16 However, most studies are small, uncontrolled, single-center case series with short-term follow-up and potential selection bias (e.g., tertiary center urology clinic recruitment). Thus, the long-term risk and timing of adverse kidney outcomes in posterior urethral valves are poorly understood, particularly after transition to adulthood. These data are needed to improve patient counseling, develop evidence-based follow-up guidelines, target interventions to improve posterior urethral valve care, and raise awareness about the need for improved transitional urologic care. To address this knowledge gap, we conducted a population-based cohort study using provincial health care administrative databases in Ontario, Canada. Our primary objective was to determine the incidence, risk, and timing of major adverse kidney events (MAKE) among children diagnosed with posterior urethral valves compared with surgical and general population controls.

Methods

Study Design

We conducted a retrospective, population-based cohort study of all male children (age <2 years) living in Ontario with Ontario Health Insurance Plan (OHIP) coverage between April 1, 1991, and March 31, 2021. Ontario is Canada's largest province, with a pediatric population of approximately 7 million during the study period, and a government-funded universal health care system. This study was authorized under section 45 of Ontario's Personal Health Information Protection Act and did not require research ethics board approval. This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology¹⁷ and REporting of studies Conducted using Observational Routinely collected health Data¹⁸ guidelines (Supplemental Appendix 1).

Data Sources

We used linked provincial health administrative databases within ICES. ICES is an independent, nonprofit research institute whose legal status under Ontario's health information privacy law allows it to collect and analyze deidentified health care and demographic data without consent for health system evaluation and improvement. Emigration from Ontario is uncommon (<1% annually) and is the sole reason for loss to follow-up.¹⁹ The Registered Persons Database contains demographic, birth, and death information. The Canadian Institute for Health Information Discharge Abstract Database, Same Day Surgery, and National Ambulatory Care Reporting System were used to identify diagnoses, comorbidities, procedures, and outcomes from inpatient, same-day surgery, and emergency department visits. The OHIP database contains physician billing records (diagnostic/procedural) for all inpatient and outpatient services. The Canadian Organ Replacement Register contains transplantation data. Laboratory data were obtained using the Ontario Laboratories Information System database. Ontario Laboratories Information System is a provincial repository of data from hospital-based and community laboratories across Ontario. The Ontario Mother-Baby database links maternal records with all hospital-based births. These databases were linked using unique encoded identifiers, and data analysis was performed at ICES Western, London, Ontario. Complete, uncleaned data were available to study investigators. Supplemental Appendix 2 includes descriptions of each database and administrative definitions used for each variable.

Study Population

We identified all male children age <2 years in Ontario between April 1, 1991, and March 31, 2021. We excluded non-Ontario residents, those ineligible for OHIP coverage, and those with congenital anomalies of the abdominal wall (prune belly, gastroschisis, omphalocele, or exomphalos). Because this was a population-based study, a formal sample size calculation was not performed.

Exposed and Comparator Cohorts

Our main exposure was posterior urethral valve diagnosis, defined using International Classification of Diseases (ICD-9 [1991–2002] and ICD-10 [2003–2021]) diagnostic codes for posterior urethral valves and OHIP procedural codes for valve ablation (S542). These diagnostic codes are specific to posterior urethral valves and have been used in prior health administrative database research.²⁰ The cohort entry date for children with posterior urethral valve was the date of the first diagnosis or procedure. We included two unmatched comparator cohorts: (1) the general population (all male children age <2 years who were not diagnosed with posterior urethral valves) and (2) children with pyeloplasty (all male children age <2 years who underwent pyeloplasty surgery [OHIP procedural code S422]). Pyeloplasty comparators represent another cohort of children treated by urologists with long-term kidney health surveillance. Children who had posterior urethral valves and underwent pyeloplasty (3%) were assigned to the posterior urethral valve cohort. The cohort entry date for pyeloplasty comparators was the procedure date. The cohort entry date for the general population was randomly assigned for the general population on the basis of the distribution of children with posterior urethral valves.

Baseline Characteristics

We evaluated the following baseline sociodemographic characteristics (age, neighborhood income quintile [by postal code],²¹ rural status [community <10,000 persons],²² and cohort entry era); maternal characteristics (maternal age, assisted reproductive technology use, maternal hypertensive disorders [gestational hypertension/pre-eclampsia], and maternal diabetes); pregnancy characteristics (oligohydramnios, fetal ultrasounds, multiple pregnancy, and prenatal interventions [amniocentesis/fetoscopy, vesicoamniotic shunt insertion, fetal body cavity aspiration (i.e., bladder), and amnioinfusion]); and birth and postnatal characteristics (gestational age, birthweight, birth asphyxia, neonatal respiratory distress syndrome, chronic lung disease, invasive ventilation, and neonatal intensive care unit [ICU] admission). We looked back from the cohort entry date for preexisting comorbidities (hypertension, CKD, acute KRT, congenital heart disease, and cardiac surgery) and health care utilization (hospitalizations, emergency department visits, and ICU admissions).

Outcomes

The primary outcome was time to MAKE, which is a composite of all-cause mortality, long-term KRT (long-term dialysis [≥2 dialysis codes separated by 90–150 days] or kidney transplant), or incident CKD, defined using validated codes.^23–26 MAKE is a patient-centered outcome, recommended for kidney outcome research,^27–29 and its components are reliably captured using health administrative databases.^{23,24,30–32} Secondary outcomes included individual MAKE components, hypertension, and AKI during hospitalization. We evaluated postnatal interventions, including posterior urethral valve ablation, vesicostomy, cutaneous ureterostomy, and gastrostomy tube insertion. All primary and secondary outcomes were ascertained using administrative data (Supplemental Appendix 2). Outcome assessment started at the cohort entry date and continued until the date of death, loss to follow-up (90 days after the end of OHIP eligibility), or March 31, 2022 (administrative censoring).

Statistical Analysis

Baseline characteristics for each cohort were reported as frequencies and proportions for categorical variables and means (SD) or medians (quartile 1–3 [Q1–3]) for continuous variables. Standardized differences were used to quantify baseline differences between cohorts. A standardized difference of ≥0.1 is considered substantial.³³ We determined the cumulative incidence and incidence rates (IRs; events per 1000 person-years) of each outcome throughout follow-up. Time-to-event analysis was performed using the Kaplan–Meier method. We used multivariable Cox proportional hazards models to determine the association (hazard ratios [HRs] and 95% confidence intervals [CIs]) between posterior urethral valves and each outcome. These models were adjusted for prespecified baseline covariates: age, income quintile, rurality, neonatal respiratory distress syndrome/chronic lung disease, neonatal ICU admission, preexisting hypertension, CKD, or cardiac surgery. Proportional hazards assumptions were checked for each model. To address proportionality assumption violations for the primary analysis, Cox models were time stratified using data-derived time periods so that the proportionality assumption was met within each period. We also used restricted cubic spline regression with four knots to model the nonlinear, time-varying HR for MAKE among children with posterior urethral valves versus general population comparators.³⁴

To evaluate for effect heterogeneity, we performed subgroup analyses using interaction terms and stratified Cox proportional hazards models to determine the association between posterior urethral valves and MAKE in prespecified subgroups: (1) full-term (≥37 weeks gestation) versus preterm delivery, (2) age at cohort entry (<90 versus 90–364 days versus 1–2 years), (3) first postnatal surgical procedure (valve ablation versus vesicostomy versus ureterostomy), and (4) cohort entry era (1991–1996, 1997–2001, 2002–2007, 2008–2014, and 2015–2020). We performed additional sensitivity analyses to test the robustness of our findings by (1) restricting our posterior urethral valves case definition to only procedural billing codes for valve ablation, (2) using children who underwent pyeloplasty as the comparator group, (3) restricting to children without preexisting CKD, and (4) defining CKD using laboratory measurements of serum creatinine and urine protein. All significance tests were two-sided, using an α=0.05 level of significance. We did not adjust the α for multiple hypothesis testing of subgroup analysis since because analyses were prespecified and exploratory. All analyses were performed at ICES using SAS version 9.4 (SAS Institute, Cary, NC).

Results

Study Population, Baseline Characteristics

Among all male children (age <2 years) living in Ontario between April 1991 and March 2021, 760 were diagnosed with posterior urethral valves, 863 underwent pyeloplasty surgery, and 1,051,530 were from the general population (Supplemental Figure 1). After exclusions, we identified 727 eligible children with posterior urethral valves, 855 pyeloplasty comparators, and 1,013,052 general population comparators (Table 1). Among children with posterior urethral valves, the median age at diagnosis was 40 days (Q1–3, 10–196), 39% were admitted to the neonatal ICU, 21% received invasive ventilation, and very few underwent prenatal intervention. Valve ablation or excision was the most common first surgical procedure (54%). Vesicostomy and cutaneous ureterostomy were performed in 151 (21%) and 23 (3%) children with posterior urethral valves, respectively. Compared with the general population, children with posterior urethral valves were more likely to be born preterm, admitted to a neonatal ICU, and invasively ventilated. Gastrostomy tube insertion was performed in 83 (11%) children with posterior urethral valves versus 5973 (1%) general population comparators.

Table 1.

Baseline characteristics among male infants who were diagnosed with posterior urethral valves, underwent pyeloplasty surgery, or were alive in Ontario between April 1991 and March 2021

Variable	Children with PUV	Pyeloplasty Comparators	Std Diffa	General Population	Std Diffa
Total patients (No.)	727	855		1,013,052
Sociodemographic characteristics
Median age at cohort entry, d (quartile 1–3)b	40 (10–196)	208 (113–355)	—	40 (12–188)	—
Income quintile, No. (%)c
1 (lowest)	194 (27)	168 (20)	0.17	240,436 (24)	0.07
2	157 (22)	170 (20)	0.04	203,456 (20)	0.04
3	145 (20)	186 (22)	0.05	208,331 (21)	0.02
4	126 (17)	189 (22)	0.12	198,342 (20)	0.06
5 (highest)	105 (14)	142 (17)	0.06	162,487 (16)	0.04
Rural status, No. (%)d	66 (9)	84 (10)	0.02	100,777 (10)	0.03
Cohort entry era, No. (%)
1991–1996	135 (19)	164 (19)	0.02	268,352 (27)	0.19
1997–2001	116 (16)	60 (7)	0.28	175,549 (17)	0.03
2002–2007	153 (21)	184 (22)	0.01	189,718 (19)	0.06
2008–2014	177 (24)	218 (26)	0.03	216,035 (21)	0.07
2015–2020	146 (20)	229 (27)	0.16	163,398 (16)	0.10
Maternal characteristicse
Median maternal age, yr (quartile 1–3)	30 (26–34)	30 (27–34)	—	30 (26–34)	—
Spontaneous conceptionf, No. (%)	267 (97)	366 (97)	0.03	253,700 (97)	0.01
Maternal hypertensive disorder of pregnancy, No. (%)	23 (4)	29 (4)	0.02	23,399 (4)	0.01
Maternal diabetes, No. (%)	30 (5)	61 (8)	0.14	34,343 (5)	0.03
Pregnancy characteristics
Oligohydramnios, No. (%)	14 (2)	0	0.21	580 (0.1)	0.20
Median number of fetal ultrasounds (quartile 1–3)	3 (1–6)	4 (2–6)	—	2 (1–3)	—
Amniocentesis/fetoscopy, No. (%)	41 (6)	34 (5)	0.08	12,745 (2)	0.22
Vesicoamniotic shunt, No. (%)	NRg	NR	—	6 (<0.1)	—
Fetal cavity aspiration, No. (%)	18 (3)	NR	—	49 (<0.1)	—
Amnioinfusion, No. (%)	NR	NR	—	92 (<0.1)	—
Birth and postnatal characteristics
Gestational age, wk, No. (%)
<32	13 (2)	11 (1)	0.04	4477 (0.4)	0.13
32 to <37	89 (12)	51 (6)	0.22	30,996 (3)	0.35
≥37	343 (47)	526 (62)	0.29	395,040 (39)	0.17
Missing	282 (39)	267 (31)	0.16	582,539 (58)	0.38
Birth weight, g, No. (%)
<1500	13 (2)	8 (0.9)	0.08	5498 (0.5)	0.12
1500 to <2500	49 (7)	46 (5)	0.05	31,940 (3)	0.16
≥2500	602 (83)	726 (85)	0.06	627,031 (62)	0.48
Missing	63 (9)	75 (9)	0.00	348,583 (34)	0.66
Birth asphyxiah, No. (%)	15 (2)	8 (1)	0.10	6306 (0.9)	0.11
Neonatal RDS or CLDh, No. (%)	141 (21)	72 (9)	0.33	51,502 (8)	0.38
NICU admissionh, No. (%)	286 (39)	206 (24)	0.33	83,870 (8)	0.78
Invasive ventilationh, No. (%)	144 (21)	52 (7)	0.43	22,898 (3)	0.56
Preexisting comorbidities (from birth to cohort entry date), No. (%)
Hypertension	9 (1)	6 (0.7)	0.05	254 (<0.1)	0.16
CKD	84 (12)	31 (4)	0.31	222 (<0.1)	0.51
Acute KRT	10 (1)	NR	—	72 (<0.1)	—
Congenital heart disease	49 (7)	44 (5)	0.07	13,215 (1)	0.28
Cardiac surgery	NR	NR	—	994 (0.1)	—
Prior health care resource utilization (from birth to cohort entry date)
Mean emergency department visits (SD)i	0.7 (2)	0.8 (2)	0.07	0.2 (0.7)	0.31
Mean hospitalizations (SD)	1.2 (1)	1.5 (1)	0.30	0.7 (0.6)	0.48
Mean ICU admissions (SD)	0.3 (0.5)	0.3 (0.6)	0.00	0.1 (0.3)	0.47

CLD, chronic lung disease; ICU, intensive care unit; NICU, neonatal intensive care unit; NR, not reported; PUV, posterior urethral valve; RDS, respiratory distress syndrome; Std Diff, standardized difference.

^aStandardized difference was used to compare the posterior urethral valve cohort with the pyeloplasty comparator group (left column) and with the general population comparator cohort (right column). Standardized differences are less sensitive to sample size than traditional hypothesis tests. They provide a measure of difference between groups with respect to a pooled SD. A standardized difference ≥0.1 is considered a meaningful difference between groups.

^bThe cohort entry date was the first date of diagnosis or procedure for the posterior urethral valve cohort, the procedure date for the pyeloplasty comparator cohort, and randomly assigned for the general population comparator cohort.

^cIncome quintile was defined as neighborhood income quintile by postal code.

^dRural status was defined as residence within a community <10,000 persons.

^eMaternal characteristics were restricted to children who could be linked to a valid maternal record in MOMBABY. The number of children included for this analysis is as follows: posterior urethral valve, n=650; pyeloplasty, n=758; and general comparator, n=647,437.

^fSpontaneous conception information was only available in MOMBABY post-2009; therefore, the denominator for the measure was further restricted (posterior urethral valve, n=275; pyeloplasty, n=379; and general comparator, n=261,433).

^gNot reported because of small cell size (<6 individuals). Small cell sizes cannot be reported because of ICES privacy policies.

^hRestricted to children with a valid linked birth hospitalization record. The number of children included for this analysis is as follows: posterior urethral valve, n=683; pyeloplasty, n=791; and general comparator, n=678,082.

ⁱEmergency department visit data were available from July 1, 2000, onward. We therefore restricted analysis of baseline emergency department utilization to individuals with a cohort entry date on or after July 1, 2000.

MAKE

The median follow-up time was 16.6 years (Q1–3, 8.6–24.5) in the overall cohort (14.2 years [children with posterior urethral valves], 12.7 years [pyeloplasty comparators], and 16.6 years [general population]). Complete follow-up data until March 2022 were available for 89% of participants. Throughout follow-up, 235 (32%) children with posterior urethral valves developed MAKE versus 50 (6%) pyeloplasty comparators and 8198 (1%) general population comparators (Table 2). The cumulative incidence of MAKE among children with posterior valves was 17% by 1 year, 24% by 5 years, and 27% by 10 years of follow-up (Figure 1). The incidence of MAKE throughout follow-up was 28.3 (95% CI, 25.0 to 32.2) events per 1000 person-years among children with posterior urethral valves, higher than among pyeloplasty (IR, 4.5; 95% CI, 3.4 to –5.9) or general population comparators (IR, 0.5; 95% CI, 0.5 to 0.5). Children with posterior urethral valves were at significantly higher risk of developing MAKE throughout follow-up versus the general population (adjusted HR, 36.6; 95% CI, 31.6 to –42.4). However, the proportionality assumption was violated. To address this, we used time-stratified Cox models with data-derived time periods. The adjusted HR for MAKE among children with posterior urethral valves versus general population comparators was 69.7 (95% CI, 56.1 to 85.7) between 0 and 1 year, 42.4 (95% CI, 30.4 to 57.4) between 1 and 5 years, and 16.5 (95% CI, 12.3 to 21.5) >5 years after cohort entry. Restricted cubic spline analysis (Figure 2) demonstrated that HRs for MAKE comparing children with posterior urethral valves versus general population comparators declined over the first 5–10 years but remained significantly elevated for >30 years of follow-up.

Table 2.

Kidney outcomes throughout follow-up among children with posterior urethral valves, pyeloplasty comparators, and general population comparators

Outcomesa	Children with PUV (n=727)		Pyeloplasty Comparators (n=855)	General Population (n=1,013,052)	HRb PUV versus General Population
	No. (%)	Incidence per 1000 PY (95% CI)	No. (%)	No. (%)	Adjusted HR	95% CI
Primary outcomes
MAKEc	235 (32)	28.4 (25.0 to 32.2)	50 (6)	8198 (1)	36.6	31.6 to 42.4
Secondary outcomes
All-cause mortality	17 (2)	1.6 (1.0 to 2.6)	NRd	4355 (0.4)	2.9	1.7 to 4.8
CKDe	198 (31)	26.2 (22.8 to 30.1)	44 (5)	4017 (0.4)	82.0	70.8 to 94.9
Kidney failuree	61 (8)	6.1 (4.7 to 7.8)	6 (0.7)	155 (<0.1)	NR	NR
Hypertension	133 (18)	14.6 (12.3 to 17.3)	48 (6)	29,445 (3)	5.7	4.7 to 6.8
AKI during subsequent hospital encounter	145 (20)	15.8 (13.4 to 18.6)	32 (4)	4513 (0.5)	27.9	22.8 to 34.2

CI, confidence interval; HR, hazard ratio; MAKE, major adverse kidney event; NR, not reported; PUV, posterior urethral valve; PY, person-years.

^aMedian follow-up time was 14.2 years for children with posterior urethral valve, 12.7 years for pyeloplasty comparators, and 16.6 years for general population comparators.

^bUsing Cox proportional hazards models for time-to-event outcome comparing children with posterior urethral valve and the general population. The proportional hazards assumption was violated in all primary and secondary outcome models. Each model was adjusted for age, income quintile, rurality, neonatal respiratory distress syndrome/chronic lung disease, neonatal intensive care unit admission, preexisting hypertension, CKD, and cardiac surgery.

^cMajor adverse kidney events was defined as a composite of all-cause mortality, long-term KRT (long-term dialysis or kidney transplant), or CKD.

^dNot reported because of small cell size (<6 individuals). Small cell sizes cannot be reported because of ICES privacy policies.

^eThe denominators for the CKD and kidney failure outcomes were restricted to individuals without preexisting CKD and kidney failure, respectively.

Figure 1.

Cumulative probability of MAKE among children diagnosed with posterior urethral valves versus general population comparators. MAKE were defined as a composite of all-cause mortality, long-term KRT (long-term dialysis or kidney transplant), or CKD. MAKE, major adverse kidney event; PUV, posterior urethral valve.

Figure 2.

Smoothed HRs for long-term MAKE among children diagnosed with posterior urethral valves versus general population comparators by restricted cubic spline analysis. MAKE were defined as a composite of all-cause mortality, long-term KRT (long-term dialysis or kidney transplant), or CKD. The solid line represents the HR for MAKE over time after cohort entry, comparing children with posterior urethral valves versus the general population. The shaded area represents the pointwise 95% CIs for the corresponding HR. The markers indicate the position of four knots (with equal number of outcome events in each interval). The dashed line represents a HR of 1 (i.e., no significant difference in the hazard of MAKE between children with posterior urethral valves and general population comparators). The HR of MAKE was significantly elevated among children with posterior urethral valves than the general population throughout follow-up. The HR of MAKE declined over the first 5 years after posterior urethral valve diagnosis but remained elevated for >30-year follow-up. The CIs should be interpreted cautiously because they are not global CIs and thus not adjusted for multiplicity. CI, confidence interval; HR, hazard ratio.

Secondary Outcomes

The cumulative incidence, IRs, and HRs for secondary outcomes are presented in Table 2. Throughout follow-up, 17 (2%) children with posterior urethral valves died versus 4355 (0.4%) in the general population. Among children with posterior urethral valves, the cumulative incidence of CKD was 31%, kidney failure was 8%, hypertension was 18%, and AKI during a subsequent hospitalization was 20%. The risk of each MAKE component (all-cause mortality, CKD, and kidney failure), hypertension, and AKI was higher among children with posterior urethral valves than either general population (Table 2) or pyeloplasty comparators (Supplemental Table 1).

Subgroup Analyses

The risk of MAKE was higher among children with posterior urethral valves diagnosed before 90 days of life versus children diagnosed later (Figure 3). Children diagnosed in a more recent era (i.e., after 2008) also had a higher risk of MAKE compared with those diagnosed in earlier eras. Children who underwent vesicostomy as their first surgical intervention were also at higher risk of MAKE than children who underwent valve ablation. There was no evidence of interaction between prematurity and MAKE among children with posterior urethral valves.

Figure 3.

Subgroup analyses for MAKE among children diagnosed with posterior urethral valves versus general population comparators. We evaluated for effect modification on the association between posterior urethral valve diagnosis and MAKE among the above subgroups using interaction terms (far right column) and stratified Cox proportional hazard models to determine HR for MAKE among children with posterior urethral valves versus general population comparator cohorts for each stratum. The proportional hazards assumption was violated in all subgroup models. The number of patients (second left column) indicates the proportion of the overall study population in each subgroup. For the first surgery subgroup, this indicates the number of children diagnosed with posterior urethral valves who underwent each procedure as their first surgery. For the first surgery subgroup analysis, the comparator cohort was divided into three equally sized cohorts (335,559 children) for comparison with each surgical intervention.

Sensitivity Analyses

Children with posterior urethral valves were at significantly higher risk of developing MAKE throughout follow-up versus pyeloplasty comparators (adjusted HR, 4.4; 95% CI, 3.2 to 6.1). After restricting our posterior urethral valve case definition to only children who underwent valve ablation, the risk of MAKE throughout follow-up remained higher (adjusted HR, 50.1; 95% CI, 41.3 to 60.9) than the general population. There was also no substantial change in the risk of MAKE (adjusted HR, 40.5; 95% CI, 35.1 to 46.6) after excluding children with preexisting CKD. Children with posterior urethral valves were also at significantly higher risk of laboratory-defined CKD, although the cumulative incidence of CKD varied based on the definition used (Supplemental Table 2).

Discussion

In this large population-based cohort study, we found that children with posterior urethral valves were at substantially higher risk of long-term MAKE compared with the general population or children who underwent pyeloplasty surgery. Over a 14.2-year median follow-up, 32% of children with posterior urethral valves developed MAKE, 31% developed CKD, and 8% developed kidney failure. The risk of MAKE declined over the first 5–10 years after diagnosis but remained elevated for >30 years. This highlights the need for kidney health surveillance among adults with a history of obstructive uropathy. Although the risk of mortality was three times higher among children with posterior urethral valves than the general population, only 2% died throughout follow-up, which is lower than previous estimates.

Children with posterior urethral valves are at risk of developing CKD and kidney failure over time. Previous case series report that 32%–54% develop CKD and 15%–23% develop kidney failure.^9–16,35 A meta-analysis of 13 studies including 1731 children with posterior urethral valves found that 38% developed CKD and 14% kidney failure over mean 5.5-year follow-up.³⁶ The risk of developing kidney failure was 5.2 times higher among 3198 Israeli children with congenital anomalies of the kidneys and urinary tract (including posterior urethral valves) than children without childhood kidney disease.³⁷ Progressive decline in kidney function is observed during childhood and adulthood.³⁸ However, there is considerable variability in these estimates on the basis of follow-up duration, study population, timing of diagnosis, interventions, and CKD definitions. Furthermore, these are primarily single-center or single-surgeon studies, which limits generalizability. Two recent multicenter studies have evaluated long-term posterior urethral valves outcomes. McLeod et al. included 274 children with posterior urethral valves who underwent surgery before 90 days of life at five US children's hospitals between 1995 and 2005.³⁹ Children were followed until a median age of 6.3 years, and 15% initiated long-term KRT. KRT risk was higher between age 10 and 15 years. Among children >13 years at last follow-up, 25% had initiated long-term KRT. Herbst et al. studied 685 children hospitalized for posterior urethral valves between 1992 and 2006 in the US Pediatric Health Information System.²⁰ Over a median 7-year follow-up, 5% died, 9% underwent dialysis catheter insertion, and 7% underwent kidney transplant. Kidney failure (dialysis catheter insertion or kidney transplant) occurred in 13% by 10-year and 20% by 15-year follow-up. Both studies used historical cohorts of children diagnosed before 2006, which is a limitation. Furthermore, both cohorts had high loss to follow-up and were restricted to children treated at tertiary children's hospitals, which introduces bias. Changes in management, follow-up care, and kidney health surveillance in recent decades may reduce the risk of long-term kidney failure.

Compared with these studies, we found a lower incidence of kidney failure (8%) over median 14.7-year follow-up. This may reflect differences in our study population; we included all Ontario children diagnosed with posterior urethral valves. Herbst et al. defined kidney failure on the basis of dialysis catheter insertion, regardless of whether dialysis was subsequently performed. This may overestimate kidney failure incidence, particularly among infants who undergo peritoneal dialysis catheter insertion but have kidney functional recovery. We found that children diagnosed with posterior urethral valves were at 37 times and four times higher risks of MAKE than children from the general population or who underwent pyeloplasty surgery, respectively. This is consistent with a study by Matsell et al. that found that posterior urethral valves were associated with a 4.7 times greater odds of CKD than other congenital anomalies of the kidneys and urinary tracts.⁴⁰

Several consistent predictors of long-term kidney outcomes have been identified in posterior urethral valves, including the nadir serum creatinine level, severity of renal dysplasia, bladder dysfunction, proteinuria, and bilateral vesicoureteral reflux.^{10,13,16,35,41,42} A machine learning model including baseline kidney function, nadir creatinine, vesicoureteral reflux grade, and renal hypodysplasia was found to have a high predictive ability (c-index 0.95) for long-term KRT.⁴³ There are conflicting data on the relationship between the timing of posterior urethral valve diagnosis and kidney outcomes.⁴⁴ Several authors have found that prenatal or early postnatal diagnosis is associated with a higher risk of subsequent CKD.^13,42,45 Earlier detection may reflect more severe disease and irreversible renal dysplasia. However, other authors report that later detection is associated with worse outcomes^16,46,47 or no difference in outcomes.^48,49 Although late diagnoses typically have a milder phenotype, missed cases of severe posterior urethral valves are likely to have progressive CKD due to bladder dysfunction and pressure-mediated kidney injury. In our study, earlier diagnosis (age <90 days) was associated with a higher risk of MAKE. We also found a graded association between more recent diagnosis era and MAKE. This may reflect increasing CKD surveillance and detection, as well as improvements in the diagnostic accuracy of administrative codes for CKD (coding creep) over time.

Previous studies report that 8%–39% of children with posterior urethral valves develop hypertension.^10–12,50 More than 50% of all children with CKD have hypertension and most have suboptimal BP control, which is associated with CKD progression.^51–55 As a result, lower BP targets are recommended for children with CKD.^56–58 In our study, 18% of children with posterior urethral valves developed hypertension, which was six times higher than the general population. This justifies regular BP screening (at least annually) of all children and adults with a history of obstructive uropathy. Early hypertension detection and treatment may slow CKD progression and prevent future cardiovascular disease.^55,59 Mortality in children with posterior urethral valves has improved considerably over time. In the 1960s, up to 40% of children born with posterior urethral valves did not survive, with death typically in the neonatal period.⁶⁰ In recent decades, <5% of children with posterior urethral valves die.^10,20,42 This reflects improvements in prenatal detection and management, earlier postnatal intervention, and improved neonatal ICU care for pulmonary hypoplasia, which is the major cause of neonatal death.²⁰ Lower mortality may also relate to more elective fetal terminations, particularly of severe cases that are at the highest risk of death.⁸ Finally, advances in pediatric KRTs may contribute to lower mortality among children with kidney failure. In our study, only 2% of children with posterior urethral valves died, but death was three times more common than among the general population.

This study has multiple strengths. We included a large, population-based, multicenter cohort of children with posterior urethral valves with up to 31 years of follow-up into early adulthood. Loss to follow-up during the study period was uncommon (approximately 10%). We quantified the long-term risks of CKD, kidney failure, and death among children with posterior urethral valves compared with the general population or children undergoing other urological surgery. Multivariable adjustment was used to address potential confounders. Nevertheless, this study has several limitations intrinsic to using health administrative databases, including misclassification of exposures and outcomes. Specific diagnostic codes exist for posterior urethral valves, which have been used in previous research.²⁰ Still, their performance has not been formally validated, and voiding cystourethrogram data were not available to confirm the diagnosis or presence of vesicoureteral reflux. The incidence of posterior urethral valves in Ontario on the basis of these diagnostic and procedural codes ranged from 1 in 3000–6000 live births, consistent with previous estimates.^4–6 Procedural claim data are reliable because they are a physician and hospital reimbursement requirement. Components of MAKE are accurately captured in health administrative databases. The positive predictive values of administrative definitions for KRT and death are >85%–96%.^24–26 Administrative coding definitions for pediatric CKD and hypertension both have high specificity but low sensitivity (CKD: specificity 98%, sensitivity 38%; hypertension: specificity 100%, sensitivity 19%).^23,61 Thus, CKD and hypertension incidence is underestimated using these administrative definitions. Although laboratory data were incomplete in Ontario health administrative databases, we found that children with posterior urethral valves had a higher risk of laboratory-defined CKD, regardless of the definition used. Higher rates of surveillance among children with posterior urethral valves may lead to earlier CKD and hypertension detection. Posterior urethral valves were also associated with a higher risk of death and kidney failure, which are less susceptible to surveillance bias. Finally, we lacked data on several important potential effect modifiers, including renal dysplasia, nadir creatinine levels, and bladder dysfunction.

In this population-based cohort study, we found that one third of children diagnosed with posterior urethral valves developed MAKE by early adulthood. Their risk of MAKE is substantially higher than children from the general population or who undergo pyeloplasty and persists well into adulthood. This justifies long-term kidney health and BP surveillance among children and adults with a history of obstructive uropathy. This may include dedicated follow-up clinics; standardized care pathways involving multidisciplinary collaboration between urologists, nephrologists, pediatricians, and allied health professionals (e.g., nurses, dietitians, pharmacists, and social workers); and further research comparing prenatal and postnatal interventions used to optimize bladder drainage, prevent urinary tract infections, and limit postnatal kidney damage. At our center, a dedicated posterior urethral valves clinic with standardized care pathways improved rates of prenatal diagnosis, early surgical intervention, medication use for bladder dysfunction, and the time to nadir creatinine level.⁶²

As more children with posterior urethral valves survive into adulthood, the transition to adult urologic care becomes key to optimizing long-term kidney outcomes.⁶³ However, there are substantial system-, physician-, and patient-related barriers to implementing effective transitional care. Future research should evaluate the impact of transition strategies, including integrated clinics and standardized communication tools.

Supplementary Material

jasn-35-1715-s001.pdf ^{(1.4MB, pdf)}

jasn-35-1715-s002.pdf ^{(437.1KB, pdf)}

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/JSN/E820.

Funding

A.J. Lorenzo: Hospital for Sick Children (Division of Urology Funding).

Author Contributions

Conceptualization: Rahul Chanchlani, Joana Dos Santos, Nivethika Jeyakumar, Armando J. Lorenzo, Mandy Rickard, Cal H. Robinson, Tim Van Mieghem.

Data curation: Nivethika Jeyakumar, Cal H. Robinson, Graham Smith.

Formal analysis: Nivethika Jeyakumar, Cal H. Robinson, Graham Smith.

Funding acquisition: Joana Dos Santos, Armando J. Lorenzo, Mandy Rickard.

Investigation: Rahul Chanchlani, Joana Dos Santos, Nivethika Jeyakumar, Armando J. Lorenzo, Juliane Richter, Mandy Rickard, Cal H. Robinson, Graham Smith, Tim Van Mieghem.

Methodology: Rahul Chanchlani, Joana Dos Santos, Armando J. Lorenzo, Mandy Rickard, Cal H. Robinson.

Project administration: Nivethika Jeyakumar, Graham Smith.

Supervision: Rahul Chanchlani, Armando J. Lorenzo.

Visualization: Cal H. Robinson.

Writing – original draft: Cal H. Robinson.

Writing – review & editing: Rahul Chanchlani, Joana Dos Santos, Nivethika Jeyakumar, Armando J. Lorenzo, Juliane Richter, Mandy Rickard, Cal H. Robinson, Graham Smith, Tim Van Mieghem.

Data Sharing Statement

Data cannot be shared. Deidentified individual participant data will not be made available based on privacy restrictions governing the use of ICES data (https://www.ices.on.ca/data-privacy/).

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/JSN/E819.

Supplemental Figure 1. Participant flow diagram.

Supplemental Table 1. Sensitivity analysis for hazard ratios for kidney outcomes throughout follow-up, comparing children with posterior urethral valves to pyeloplasty controls.

Supplemental Table 2. Sensitivity analysis for cumulative incidence of CKD using laboratory-based definitions, among children with posterior urethral valves and the general pediatric population.

Supplemental Appendix 1. The RECORD statement, checklist of items, extended from the STROBE statement, which should be reported in observational studies using routinely collected health data.

Supplemental Appendix 2. Administrative codes used for cohort selection, exposures, baseline characteristics, and outcomes.