Changes in the management of urinary tract infections (UTIs) and patients with morphological abnormalities of the urinary system have been very uneven over time. Some have resulted from technical innovations, such as the introduction of prenatal or neonatal ultrasound examination of the urinary system. Until the introduction of ultrasound, the standard of care following a first UTI in children was routine performance of urethrography and cystography.1 The advent of sonography brought on a partial change to existing imaging protocols, introducing the factor of age (urethrography and cystography in infants aged less than one year and ultrasound examination combined with radionuclide cystography from age 1 year).2 Additional changes according to the sex of the patient were introduced soon after.3

In 2008, Isky Gordon et al. coined the term “high-risk” children in this context, that is, children at increased risk of vesicoureteral reflux (VUR), who would constitute the group in whom radiology or scintigraphy techniques would be indicated based on the findings of the ultrasound examination performed after an episode of febrile UTI. This introduced a new notion, that is, that there were children at “low risk” for VUR in whom, therefore, it would not be necessary to make extensive imaging assessments to screen for low-grade VUR.4 In recent years, efforts have been made to establish criteria for the rational use of cystography,5 especially since the long-held belief that UTIs could cause chronic kidney disease in the absence of renal morphological abnormalities has been dismissed.6

Past and recent recommendations from experts and working groups omit kidney function testing for ordering a cystogram, except for a recent mention in the Spanish Clinical Practice Guidelines on Urinary Tract Infections in the Pediatric Population.7 In this context, we conducted a study to assess the importance of performing tests to assess kidney function, especially in relation to fluid and electrolyte balance, to determine the indication of cystography in the pediatric population.

This study included both patients with morphological abnormalities detected in the prenatal or neonatal ultrasound as well as those who had had one or more episodes of UTI. These are the two main reasons why cystography is usually ordered in everyday clinical practice.

In agreement with what has been known for some time, our series showed a clear association between UTI and VUR. However, it should be noted that half of the UTIs were diagnosed in the absence of VUR.

We did not find an evident association between VUR and renal pelvis dilation with an APD of less than 2 cm (Tables 1 and 3). It is known that mild/moderate dilation is not sufficiently sensitive to suspect VUR,13,14 even applying the new, currently used system for the classification of what used to be known as renal ectasia (Urinary Tract Dilation [UTD] classification system).14,15 In consequence, we did not include cases of renal pelvis dilation with an APD of less than 2 cm in the SMA subgroup. The recommendation in past articles to perform cystography in every case of urinary tract dilation appears to be obsolete.16

We propose the concept of SMA confirmed by renal ultrasound or scintigraphy (DMSA scan) (Table 1). The prevalence of SMAs increased in relation to the severity of VUR, from 28.6% in cases of mild VUR to 100% in cases of grade V VUR (Table 3). The presence of these anomalies was significantly associated with the presence of VUR (OR, 5.91; P < .0001).

With regard to renal function, we did not measure laboratory markers that require blood collection to assess glomerular renal function, such as creatinine or cystatin C. These markers are highly specific for kidney injury but offer a very low sensitivity, and by the time abnormal values are found, the reduction in functioning renal parenchyma is considerable. We measured three functional markers of demonstrated utility for early detection of kidney injury that are also easy to measure in spot urine samples, the maximum UOsm and the albumin/creatinine and NAG/creatinine ratios. The last two were not significantly associated with the presence of VUR in our sample. On the other hand, we found normal maximum UOsm values in 61% of patients without evidence of VUR and abnormal values in 75% of patients with VUR (43/57) (Table 2). None of the children with severe VUR (grades IV and V) had a maximum UOsm in the normal range (Table 3). However, of the cases of moderate VUR, only 52% were associated with urine concentrating defects (Table 3), which constitutes a partial limitation to our conclusions, as some antibiotic prophylaxis could be beneficial in some of these cases depending on the circumstances.

The sensitivity of the urine concentration test for detection of kidney injury is higher compared to other markers because it is a global marker indicative of tubular integrity when its results are normal, that is, it is not a marker of a specific tubule segment, but of the tubules overall, as normal results depend on the preserved handling of sodium and chloride in each functional segment, preserved structure and function of the medullary countercurrent mechanism, and normal water permeability of the tubules in response to vasopressin.10,12

The pathophysiological mechanism by which VUR results in a reduced maximum UOsm involves the additive effects of the increased pressure exerted on the renal parenchyma and the associated loss of nephron mass, which may be congenital or acquired.17,18

Our findings as well as previously published data demonstrate that changes in renal water handling are directly associated with the grade and severity of VUR.10,19,20 Thus, the effects of increased pressure on renal function have been investigated in experimental models by producing ureteral obstruction, which translates to a reduction in the activity of some tubular sodium channels,21,22 in the expression of urea transporters23 and the activity of aquaporins expressed in the collecting ducts.24 Specifically, one experimental study demonstrated the negative impact of VUR on the expression of aquaporins 1 and 2.25

Table 7 offers a summary of the articles published by our group, presenting the evidence gathered on the efficacy of basic renal function tests (maximum UOsm, albumin/creatinine and NAG/creatinine ratios) in common conditions characterized by the loss of renal parenchyma or increased pressure on the urinary tract.10,12,26–30 As can be seen in the table, the urine concentrating capacity is the most sensitive marker of functional renal impairment in various conditions commonly encountered in pediatric care. The key issue investigated in the present, complementary study is whether the maximum UOsm can be a useful marker for the purpose of determining the indication of cystography.

Table 8 presents the indications for cystography published in recent years in various guidelines and articles.7,31–36 The broader indications applied decades ago have been abandoned1–3 and more specific indications are being pursued, chiefly aiming at the detection of high-grade VUR.4,33 Recent criteria, while heterogeneous, tend to agree on the performance of cystography in the presence of severe morphological abnormalities detected by sonography or DMSA scintigraphy, in young male patients and in the case of recurrent febrile UTIs, especially when caused by bacteria other than Escherichia coli.7,31–36 It is widely accepted that isolated renal pelvis dilation is not an indication for performing imaging tests beyond the ultrasound.32

It is remarkable that, both in the past and today, criteria such as age or clinical factors are taken into account to determine the indication for cystography, given the disparity and potential for confusion that this may entail in the selection of candidates, for example, in the case of a girl aged more than 2 or 3 years of age with a febrile UTI or cystitis caused by a bacterium other than E coli. However, the presence of morphological abnormalities is an identifiable and measurable criterion that is not as variable.

Table 8 reflects the omission of renal function testing in all articles, save for a passing and inexplicit mention in the recent Spanish guideline.7

The three predictor combinations presented in Table 5 had NPVs around 90%, that is, performed well as criteria to determine the indication of cystography that could prevent some unnecessary tests. The LR− obtained for isolated variables (0.53 for morphological abnormalities and 0.4 for the maximum UOsm) were superior to those obtained for the three combinations (0.22 in two of them and 0.15 in the other), although these values are considered moderate changes in the probability of ruling out VUR. The PPV of the three combinations ranged between 41.5% and 49%. The reason for this is that both morphological abnormalities on imaging and VUR can be present in the absence of a history of UTI, especially in children who have received antibiotic prophylaxis. Thus, the association of impaired urine concentrating capacity with morphological abnormalities or a relevant clinical history (UTI) improves diagnostic accuracy (with increases in the NPV and the odds ratio and a decrease in the LR−), although with a lesser specificity in the latter case which, therefore, would result in the performance of cystography in a larger number of patients.

On the other hand, many cases of mild or moderate urinary tract dilation that do not result in obstruction or are not associated with VUR do present with urine concentrating defects (Table 2). No hypothesis has been proposed as to the underlying mechanism of this association, although it may involve the functional immaturity of the renal medulla, with or without the presence of abnormally short loops of Henle. A longitudinal study conducted by our group identified a trend of progressive normalization over time.37

Applying the proposed eligibility criteria, none of the cases of severe VUR (grade IV-V) would have been missed if the maximum UOsm was included (Table 6). The combination “SMAs or reduced maximum UOsm” clearly simplifies the initial indication of cystography without needing to consider age, sex, or the location or causative agent of UTI. The other advantage of this simple criterion, as we already noted, is that it does not fail to identify any patients with severe VUR, who are ultimately those who require closer monitoring and possibly surgical intervention.35,38

Our study has the limitations intrinsic to its retrospective design, such as potential variation in the diagnosis of UTI, the fact that urinary albumin and NAG were not measured in all patients or the greater difficulty in interpreting the findings of contrast‑enhanced voiding urosonography for reflux grading. In addition, also due to its retrospective nature, the fact that our group has been convinced for many years of the utility of measuring UOsm in everyday practice may have resulted in the decision not to perform cystography in patients with a normal urine concentrating capacity, which could have affected our estimation of diagnostic accuracy.

In short, based on our findings, if SMAs are present in children with a previous history of UTI, measurement of the maximum UOsm would not be necessary to determine the indication of cystography. However, measurement of UOsm would slightly improve the NPV without missing any cases of grade IV VUR. In cases in which severe VUR needs to be ruled out in a patient without a history of UTI, a urine concentration test would be helpful. One of the drawbacks of the functional test that we propose is that urine concentrating capacity is preserved in nearly half of the patients with moderate VUR, most likely because the increase in pressure is not significant, and, as we noted above, some of these patients could benefit from antibiotic prophylaxis. In any case, the performance of renal tubular function tests as a complement to other clinical, laboratory and sonographic variables can help guide decision-making and contribute to preventing the performance of unnecessary cystograms.