Pyuria is a useful marker for assessing urinary tract infection (UTI) in the general population. The presence of pyuria is, in general, highly suggestive of UTI, especially in symptomatic patients. Pyuria seems to be common in patients with chronic kidney disease (CKD). Sterile pyuria is considered to occur in CKD owing to chronic renal parenchymal inflammation [1^,2]. However, there are limited data on whether CKD increases the rate of pyuria or how pyuria in CKD should be interpreted [2]. In fact, to our knowledge, the prevalence or characteristics of pyuria has never been investigated, especially in non-dialysis CKD patients.

Studies have addressed the diagnostic value of pyuria in CKD patients, particularly those undergoing dialysis. The prevalence of pyuria in dialysis patients has been reported to range from 31% to 72% [1^,2^,3^,4^,5^,6], and the rate of documented UTI in these patients ranges from 25% to 45%. In addition, the presence of pyuria may not indicate UTI in dialysis patients [1^,4^,5^,7]. Cellular analysis of ascites is useful for early diagnosis of spontaneous bacterial peritonitis in liver cirrhosis patients, and induced sputum analysis provides information on the inflammatory phenotype of airway disease [8^,9]. In the early 1970s, Cabaluna, et al. [1] reported that pyuria is a common finding in hemodialysis (HD) patients that often occurs in the absence of infection. Of interest was the observation that the majority of white blood cells (WBCs) in the urine seemed to be lymphocytes. Unfortunately, no studies have explored urinary cell analysis as a follow-up to these observations of Cabaluna, et al. [1].

We examined whether CKD patients have an increased prevalence of asymptomatic pyuria (ASP) compared with the non-CKD population. In addition, we analyzed urinary WBCs to investigate the characteristics of pyuria and the usefulness of urinary cell analysis in identifying sterile pyuria in the CKD population.

The CKD population had higher ASP prevalence (in both male and female patients) than the non-CKD population, and it increased with CKD stage. Pyuria was more common in female patients and patients with diabetes, higher stages of CKD, proteinuria, hematuria, lower Hb and Alb, and higher CRP. Over 70% of the pyuria cases were sterile, and the majority of urinary WBCs were neutrophils, even in sterile pyuria. However, the percentage of neutrophils was significantly lower in sterile pyuria group than in UTI group (70.2% vs. 92.8%). In multivariate logistic regression analysis, the degree of pyuria, percentage of neutrophils, and presence of urinary nitrites were independently associated with sterile pyuria.

In general, CKD patients present with many comorbidities, including diabetes and neurogenic bladder that contribute to bladder stasis, thereby increasing the risk of UTI. Data on the prevalence of UTI among CKD patients is limited, although some authors have suggested that the frequency of UTI might not differ from that in the general population [14]. In the present study, the prevalence of bacteriuria was 6.6% (1.1% in males and 14.3% in females) in the asymptomatic CKD population. We were unable to determine the prevalence of asymptomatic bacteriuria (ASB) in the general population because the regular health examinations did not include a urine culture. A recent meta-analysis showed that diabetic patients have a higher prevalence of ASB than healthy controls (12.2% vs. 4.5%) [15].

The diagnosis of UTI in CKD patients can be challenging. UTI symptoms are mostly related to voiding, which is particularly reduced in end-stage renal disease patients. Moreover, it has been suggested that sterile pyuria is common in CKD patients [1^,16]. Our results confirm that ASP is more common in the CKD than in the non-CKD population (24.1% in non-dialysis CKD and 51.4% in HD patients vs. 7.2% in the non-CKD population). A previous study reported a significantly higher ASP prevalence in both male (12.2% vs. 3.4%) and female (21.4% vs. 8.7%) diabetic patients compared with that in the non-diabetic population [17]. In the present study, ASP was even more common in the CKD population (17.2% in males and 47.5% in females) than in the diabetic population. The factors underlying higher ASP prevalence in the CKD population, despite the relatively modest bacteriuria prevalence, are unknown and need to be further examined.

It is noteworthy that 78% of the pyuria cases were sterile. Overall, 24.2% of the asymptomatic CKD population (17.3% males and 34.5% females) showed sterile pyuria (Supplemental Data Table S2). A previous study reported that 2.6% males and 13.9% females have sterile pyuria in the general population; these values were lower than those observed in the CKD patients in this study [18^,19]. Moreover, the prevalence of sterile pyuria increased with increasing CKD stage (Supplemental Data Table S2). Although we could not fully determine the causes of sterile pyuria in CKD patients, various possibilities exist, including leakage of cells due to tubulointerstital injury.

While it seems that pyuria is common in CKD patients and a low degree of pyuria might be sterile, optimal cut-off values for urinary WBC counts for predicting sterile pyuria in CKD patients have not yet been established. In the present study, pyuria with ≥10 WBCs/HPF was indicative of UTI with high sensitivity (89.5%) but low specificity (53.7%), while pyuria with ≥20 WBCs/HPF showed favorable sensitivity (79.0%) and specificity (70.2%) for UTI detection. The negative predictive value of urinary WBCs at a cut-off value of 10 WBCs/HPF was 94.7%. In other words, if urinary WBC count was <10 WBCs/HPF in asymptomatic CKD patients, the possibility of sterile pyuria is >90%.

In contrast to our expectation that lymphocytes would account for most of the WBCs, the majority of urinary WBCs were neutrophils, even in sterile pyuria. This finding also contrasted with that of Cabaluna, et al. [1], although they did not report concrete differential cell count results. In this study, a low percentage of urinary neutrophils (<70%) was associated with a high probability of sterile pyuria (specificity 94.4%). The present findings are valuable as they present an additional tool for predicting sterile pyuria in pyuric CKD patients. Urinary cell analysis may help reduce unnecessary antibiotic administration or suspicion of other factors in the initial stages of asymptomatic infection. Examination of pyuria patterns through urinary cell analysis in more diverse CKD populations might provide valuable information regarding the pathobiological pathways involved in CKD.

This study had several limitations. First, despite the relatively large sample size, this was a single-center study. Second, the criteria for diagnosing UTI vary from study to study. Given that this study included an asymptomatic CKD population, UTI may actually mean ASB. However, originally, the term urinary tract infection encompassed various clinical entities including ASB, cystitis, and pyelonephritis [20]. Indeed, much of the literature concerning UTI does not differentiate between UTI and ASB [20]. Considering that urinary symptoms might be vague in CKD patients, some of the significant bacteriuria accompanying pyuria in this study might have approached clinical infection. Third, the analyses were conducted using a single urine sample from each patient, rather than using a longitudinal follow-up with multiple samples [21]. In addition, we did not examine the presence of urinary tuberculosis (TB), although it is one of the known causes of sterile pyuria. However, we considered the incidence of asymptomatic urinary TB to be low enough to dismiss the possibility.

Despite these limitations, we believe our results are important as we addressed the prevalence of sterile pyuria in an asymptomatic CKD population. To our knowledge, the prevalence of ASP or sterile pyuria has never been investigated according to CKD stage, and this is the first study to perform urinary cell analysis in pyuric CKD patients.

In conclusion, sterile pyuria is common in CKD patients. The number and distribution of urinary WBCs may be helpful for predicting sterile pyuria in CKD patients with pyuria. A large, prospective study with longitudinal follow-up may further clarify the interpretation of urinalysis in the CKD population.

The prevalence of pyuria according to the level of renal function. The prevalence of pyuria was investigated in asymptomatic CKD population. CKD was divided into HD and non-dialysis CKD; non-dialysis CKD was further classified as stages 3, 4, and 5 according to the KDIGO classification [11]. The prevalence of pyuria in the non-CKD population was estimated from regular health examination data at the study hospital and included control patients aged >40 years with eGFR ≥60 mL/min/1.73 m².

Abbreviations: CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; HD, hemodialysis; KDIGO, Kidney Disease Improving Global Outcomes.

ROC curves showing the effectiveness of the number and distribution of urinary WBCs in UTI identification. The number of WBCs was analyzed at different cut-off values: ≥10, ≥20, and ≥50 WBCs/HPF.

Abbreviations: ROC, receiver operating characteristic; UTI, urinary tract infection; WBC, white blood cells; HPF, high-power field.