Chronic asymptomatic pyuria precedes overt urinary tract infection and deterioration of renal function in autosomal dominant polycystic kidney disease

This study was performed as a retrospective, single-center, case–control study. Among the patients who registered at the ADPKD clinic at Seoul National University Hospital from Aug 1999 through Aug 2010, we retrospectively reviewed medical records of 311 individuals and collected data from 256 adult patients who received ≥ 2 urinalysis tests during their follow-up period. Patients were seen in the clinic based on their renal function: every 3-6 months for those in CKD stage I-II, every 2-3 months for those in CKD stage III, and every 6-8 weeks for those in CKD stage IV. In addition to the disease severity, we followed up patients more frequently after overt UTI episodes or when they needed acute managements such as high blood pressure or new onset hematuria. Patients with either an estimated glomerular filtration rate (eGFR) of < 15 ml/min/1.73m² or receiving renal replacement therapy were excluded from the analysis (n = 17). Patients with a brief follow-up duration of < 6 months (n = 38) were also excluded from the analysis.

ADPKD was diagnosed according to the Unified Criteria for Ultrasonographic Diagnosis of ADPKD as proposed by Pei et al [11]. We defined asymptomatic pyuria as higher than 5-9 white blood cells/high-power field (WBC/HPF) in a random urine sample without symptoms related to overt UTI. The duration of pyuria episode was defined from detection to resolution of pyuria in a subsequent urinalysis.

We categorized the patients into following 4 groups depending on the duration and frequency of the asymptomatic pyuria: no pyuria, transient pyuria, recurrent pyuria, and persistent pyuria. The no pyuria group included patients who did not experience pyuria during the study period. The transient pyuria group included patients who had fewer than 3 episodes of pyuria and patients who had ≥ 3 episodes of pyuria with inter-episode intervals longer than 6 months. Recurrent pyuria was defined as ≥ 3 episodes of pyuria within 6 months. The persistent pyuria group was defined as patients who had pyuria for ≥ 1 month. After this classification, the no pyuria and transient pyuria groups were classified as group A, and the recurrent pyuria and persistent pyuria groups were classified as group B (the chronic pyuria group).

To evaluate the effect of pyuria on the occurrence of subsequent UTI or renal function decline, we additionally divided the patients into two groups based on the number of pyuria episodes in the first year of follow up. Group^{no pyuria/1st year} was defined as the group who did not experience any pyuria episodes, and Group^{1-4 pyuria/1st year} was the sum of patients who experienced 1-4 pyuria episodes during the first year of follow up.

Acute cystitis was diagnosed based on the presence of lower urinary tract symptoms such as dysuria, a burning sensation upon voiding, frequency, urgency, or suprapubic pain with no clinical evidence of an upper urinary tract infection. Upper UTI included both APN and renal cyst infection. The APN was diagnosed clinically when patients develop rapid urinary symptoms with high fever and flank pain and the urinalysis shows signs of urinary tract infection [12]. Cyst infection was confirmed if a cyst aspirate shows definite microorganism or neutrophils debris. In addition, cyst infection was clinically suspected when fever >38.5°C develops without definite pyuria, focal tenderness over the suspected cyst, positive blood culture without urine culture positivity [13]. The imaging tools such as magnetic resonance imaging, ultrasonography, computed tomography, or positron emission tomography were taken to supplement the diagnosis of cyst infection.

Serum creatinine (sCr) was measured using the Jaffe method [14]. The Chronic Kidney Disease Epidemiology (CKD-EPI) equation was used to calculate eGFR [15, 16]. The onset of ESRD was defined as the time at which renal replacement therapy (hemodialysis, peritoneal dialysis, or transplantation) was initiated [17].

All statistical analyses were performed using SPSS version 17.0 (SPSS Inc., Chicago, IL, USA). Continuous variables were expressed as the mean ± standard deviation. The independent t-test or Mann–Whitney U test was used to compare the continuous variables among the groups, and the Chi-square test was used to analyze the categorical variables. The one-way analysis of variance (ANOVA) was used to compare the continuous variables among the 3 groups. The log rank test was used to compare UTI occurrence between group A and group B, and the multiple Cox regression model was used to identify risk factors. Differences with P < 0.05 were considered to be statistically significant.

This study was approved by the Institutional Review Board of Seoul National University Hospital (H-0901-046-269). Informed consent was obtained from the subjects in accordance with the Declaration of Helsinki.

The baseline characteristics of the 256 patients are presented in Table 1. The age of the study participants was 48.1 ± 12.8 years, and almost the same numbers of male and female subjects were enrolled. The mean follow-up duration was 65.3 ± 43.2 months, and the initial eGFR was 91.1 ± 29.2 ml/min/1.73m². Asymptomatic pyuria was observed in 26.2% of the patients at their first office visit. The incidence of asymptomatic pyuria was 0.492 episodes/patient/year. One hundred and seventy-six patients (68.8%) experienced ≥ 1 episode of asymptomatic pyuria during the course of their follow-up. Asymptomatic pyuria accounted for 93.2% of the total number of infection episodes (Table 2). The patients in group B (n = 94, 36.7%) comprised a higher percentage of females (58.5% vs. 42.0% for group A; P = 0.01) and had a longer follow-up duration (88.5 ± 37.8 vs. 51.8 ± 40.4 months for group A; P < 0.001). The prevalence of hypertension, diabetes mellitus, and urinary stones did not differ between the groups (Table 1, P > 0.05).

Urine culture was performed in only 4.6% of the patients with asymptomatic pyuria. In most of asymptomatic pyuria cases (95.4%), urine culture was not performed. Among those who underwent urine culture study, 6.5% showed negative results. The most common microorganisms were Escherichia coli (34.5%), followed by Klebsiella pneumoniae (13.8%), Staphylococcus epidermidis (11.6%), Streptococcus viridans group (9.3%), Pseudomonas aeruginosa (7.0%), Streptococcus agalactiae (7.0%), Serratia marcescens (3.4%), Corynebacterium species (3.4%), and Staphylococcus aureus (3.4%).

During the observational period, 33 (12.9%) patients developed an overt UTI. The follow-up duration was longer for the overt UTI group than for the patients without an overt UTI (81.6 months vs. 62.4 months, respectively; P = 0.01). More patients in group B experienced an overt UTI during their follow-up than the patients in group A (66.7% vs. 32.3%, respectively; P < 0.001; Table 3). Age, gender, diabetes mellitus, hypertension, and urinary stones did not increase the risk of developing an overt UTI. Compared to group A, group B had shorter periods of overt UTI-free survival and shorter periods of upper UTI-free survival (Figure 1). In addition, when we used prospective groups (Group^{no pyuria/1st year} and Group^{1-4 pyuria/1st year}) based on the number of pyuria episodes in the first year of follow up, the Group^{1-4 pyuria/1st year} demonstrated the even shorter periods of both overt UTI-free survival and upper UTI-free survival (Additional file 1: Figure S1). A Cox regression analysis was performed to determine whether group B is an independent factor for overt UTI and upper UTI. The forward stepwise method was used, and age, gender, hypertension, urinary stones, group B, and initial eGFR were included in the final analysis. In the univariate analysis, group B and lower initial eGFR were the factors associated with overt UTI. Multiple Cox regression analysis revealed that group B is an independent factor for overt UTI (hazard ratio 4.636, 95% confidence interval 1.898 to 11.323; P = 0.001; Table 4). In addition, group B independently increased the risk of upper UTI compared to group A (hazard ratio 4.612, 95% confidence interval 1.735 to 12.258; P = 0.002; Table 5). Because the incidence of acute cystitis was low, a Cox regression analysis for acute cystitis was not performed.

Among 33 patients with an overt UTI, 4 experienced both acute cystitis and an upper UTI, 6 had only acute cystitis and 23 had only upper UTI. The incidence of acute cystitis was 0.012 episodes/patient/year, and the incidence of upper UTI was 0.023 episodes/patient/year. A total of 27 patients experienced 33 episodes of upper UTI (13 episodes of APN and 20 episodes of renal cyst infection). The most common pathogen associated with upper UTI was E. coli (24.2%), followed by K. pneumoniae (9.1%). One patient had a renal abscess without a cyst infection, and this was categorized as APN. All four patients (14.8%) who experienced repeated upper UTI episodes were in group B.

The initial eGFR at the first office visit did not differ between group A and group B. However, the final eGFR in group B was significantly lower than that in group A (63.3 ± 37.0 vs. 85.5 ± 31.7 ml/min/1.73m², respectively; P < 0.001). Because the follow-up duration can affect the final eGFR value, we compared the annual change in eGFR (ΔeGFR) between group A and group B and found that group B showed a faster decline in eGFR than group A (-2.7 ± 4.56 vs. -1.17 ± 5.8 ml/min/1.73m² per year, respectively; P = 0.01; Figure 2). When we performed a subgroup analysis in baseline CKD stage I-II, group B showed even greater annual decline of eGFR compared to group A in CKD (-3.36 ± 3.47 vs. -1 ± 5.97 ml/min/1.73m² per year, P = 0.002, Additional file 2: Figure S2). However, when we performed the independent t-test in the prospective groups based on the number of pyuria episodes in the first year (Group^{no pyuria/1st year} and Group^{1-4 pyuria/1st year}), the difference in GFR decline rate disappeared (Additional file 3: Table S1 and Additional file 4: Table S2).

To analyze the risk factors that are associated with ΔeGFR, we performed a linear regression analysis. In a univariate analysis, age, diabetes mellitus, hypertension, overt UTI, and group B were associated with ΔeGFR. In a multiple linear regression model, group B was revealed as an independent risk factor for a faster decline in eGFR (B = -1.53, 95% confidence interval -2.904 to -0.155; P = 0.03; Table 6). More patients in group B (n = 15, 16.0%) reached ESRD during observational period compared to the patients in group A (n = 7, 4.3%; P = 0.001; Figure 2).

To control for the possible confounding effects of overt UTI in group B, we excluded patients who experienced overt UTI from both group A and group B. The subsequent cohorts were renamed group A^UTI- and group B^UTI- _, respectively. The patients who experienced overt UTI were separately classified as the overt UTI group irrespective of whether they were in group A or group B.

A total of 151 patients in group A^UTI-, 72 patients in group B^UTI-, and 33 patients in the overt UTI group were included in the analysis. The group B^UTI- patients exhibited a greater ΔeGFR compared to group A^UTI- (-2.6 vs. -0.9 ml/min/1.73m² per year, respectively; P = 0.009; Figure 3). The overt UTI group patients exhibited an even greater ΔeGFR at 1 year after their UTI episode (-3.6 ml/min/1.73m² per year). The clinical characteristics, including ΔeGFR, were compared between the groups (Table 7). There were statistically significant differences in gender, follow-up duration, final eGFR, ΔeGFR, and the incidence of ESRD between group A^UTI- and other groups. However, the differences between group B^UTI- and the overt UTI group were not statistically significant.