Objective To assess the effect of the duration of fever after the initiation of treatment (FAT) of febrile urinary tract infections (UTI) on the development of permanent renal lesions based on dimercaptosuccinic acid (DMSA) scintigraphy findings. To evaluate the FAT contribution to permanent renal lesion formation in relation to fever before treatment initiation (FBT), the presence of vesicourinary reflux (VUR), age and severity of infection.
Methods The inpatient records of 148 children (median age: 2.4 months (11 days to 24 months)) with a first episode of UTI during a 3-year period were analysed. DMSA findings, and clinical and laboratory parameters were evaluated.
Results Among the study population, 34/148 (22.97%) children had permanent renal lesions on the DMSA scan 6 months after a single episode of UTI. Twenty-three children (15.5%) had mild, 10 (6.7%) had moderate and 1 (0.6%) child had severe lesions on the DMSA. FAT prolongation >/48 hours was associated with older age (p=0.01) and increased absolute neutrophil count (p=0.042). The likelihood of lesions was significantly increased when FAT was ≥48 hours (R2=0.043, p=0.021). On multiple regression analysis, with the addition of FBT>/72 hours (0.022), the presence of VUR (p<0.001), C-reactive protein (p=0.027) and age (p=0.031), the effect of FAT on lesion development disappeared (p=0.15).
Conclusions Prolongation of FAT≥48 hours of febrile UTI in children <2 years significantly contributes to the development of permanent renal lesions. However, delay in treatment initiation >/72 hours, the presence of VUR, older age and infection severity seem to be more significant predictors of the development of renal lesions.
- acute pyelonephritis
- renal scarring
- fever after treatment initiation
- fever before treatment
Statistics from Altmetric.com
What is already known on this topic?
Prolongation of fever after treatment initiation in paediatric urinary tract infection may contribute to development of permanent renal lesions through the effects of proinflammatory cytokines.
Limited studies on the duration of fever after treatment initiation had conflicting results on the development of acute renal lesions.
Delayed treatment initiation, vesicoureteral reflux, older age and infection severity have been associated to the development of permanent renal lesions in children aged <2 years.
What this study adds?
Fever for ≥48 hours after treatment initiation of first urinary tract infection in children ≤2 years old was associated with the development of permanent renal lesions.
Permanent renal lesions were more severe when fever after treatment initiation lasted 48 hours or more.
However, the effect of prolonged fever >/72 hours before treatment initiation, the presence of vesicoureteral reflux, older patients’ age and severity of infection seem to be more significant predictors of the development of permanent renal lesions than fever prolongation >/48 hours after treatment initiation.
Febrile urinary tract infection (UTI) is the most common serious bacterial infection in childhood; if untreated, it can lead to permanent renal lesions as a consequence of acute inflammation of the renal parenchyma. Dimercaptosuccinic acid (DMSA) scintigraphy is the gold standard for the definition of renal involvement in febrile UTI.1–6
The importance of the duration of fever before treatment initiation (FBT) has been previously studied with controversial results.2 5 7–15 However, it is generally accepted that delaying treatment for over 48 or 72 hours can lead to acute renal lesions and consequent permanent renal lesions based on DMSA scintigraphy findings.9 10 13 15
Although the delay in treatment initiation for febrile UTI has been widely studied, the association of permanent renal lesions with the duration of fever after treatment initiation (FAT) has not. Improvement of the general clinical condition and the reduction of fever within 48–72 hours are indicative of good response to antibiotic treatment.16 17 Increased FAT is suggestive of inadequate response to treatment for several reasons, such as antibiotic resistance, host defence defects, high virulence of the pathogen and may also be affected by the delay of treatment initiation. Prolongation of fever physiologically reflects increased duration of inflammation that may contribute to tissue damage and the development of permanent renal lesions.12–14
To our knowledge, there are very limited studies on FAT and its effect on the development of permanent renal lesions.7 9 10 Thus, the aim of this study was to evaluate the relationship of the duration of FAT of febrile UTIs with permanent renal lesions based on DMSA scintigraphy findings. Another aim was to estimate the role of FAT on permanent renal lesion formation in relation to other contributing factors, such as FBT, the presence of vesicourinary reflux (VUR), age and severity of infection.
Patients and methods
Τhe inpatient records of 290 infants and toddlers with a first episode of a febrile UTI, aged 0–2 years, admitted to the Second Department of Pediatrics of the National and Kapodistrian University of Athens Medical School between 1 January 2002 and 31 December 2004 were retrospectively studied. Out of these, 148 children (median age: 2.4 months (range 11 days to 24 months, lower and upper quartiles 36 days to 7.2 months)) met the inclusion criteria and were enrolled in the present study. The inclusion criteria were: (1) ages 0–24 months; (2) fever with evidence of a UTI based on compatible symptoms and a positive urine culture (>105 colony forming units/mL for midstream clean-catch collection, >104 for catheterisation, >103 for suprapubic collection); (3) laboratory investigations including a urinalysis, a complete blood count and C-reactive protein (CRP); and (4) renal imaging studies including a renal ultrasound and voiding cystourethrogram (VCUG) in the acute phase of the infection and a 99mTc-DMSA renal scan 6 months after the infection. This scan is also called repeat DMSA scan as opposed to the initial DMSA scan (acute DMSA) that was performed a few days after the febrile UTI. The diagnosis of febrile UTI was based on the results of the urine culture and not of the urinalysis. As already reported in one of our previous articles, records of children who had a previous UTI, had received any antibiotics prior to admission or had obstructive uropathy were excluded from the study.18
Detailed medical history, physical examination, clinical features and laboratory parameters were evaluated in all patients. Fever was defined as an axillary temperature ≥38°C. FAT was divided in <48 hours or >/48 hours and was studied in relation to the development of renal scars, detected with DMSA. All repeat DMSA scans were compared with the acute DMSA scans of the patients to assess the development of renal lesions and rule out the possibility of congenital renal scars.
Urine specimens were collected by suprapubic aspiration, catheterisation or ‘midstream clean-catches’, the latter being considered positive when combined with the presence of white cells and/or bacteria in the microscopy of urine. Abnormal urinalysis was not necessary for specimens collected by suprapubic collection and catheterisation. Laboratory indices including white cell count, absolute neutrophil count (ANC) and CRP levels were measured on admission and 2 days after treatment initiation. A whole blood autoanalyser (CellTac A, Nihon Kohden, Tokyo, Japan) was used for the estimation of white cells and ANC. The serum concentration of CRP was measured by particle-enhanced immunonephelometric assays (BN ProSpect nephelometer, Dade Behring, Liederbach, Germany).
The most frequently identified pathogen in urine cultures was Escherichia coli (87.8%), followed by Proteus (4.0%), Klebsiella (4.0%), Enterobacter (2.7%) and Pseudomonas (1.3%). All patients were initially started on empiric antimicrobial therapy that was subsequently tailored according to the antibiogram. Antibiotic treatment was given intravenously, or initially intravenously and then orally, or exclusively orally depending on the age and the clinical condition of the patients. The length of treatment with appropriate antibiotics was 7–14 days according to the clinical condition. In the order of frequency administered the intravenous antibiotics were: ampicillin combined with aminoglycoside, second and third-generation cephalosporins, ampicillin, aminoglycoside, amoxicillin-clavulanate, third-generation cephalosporin combined with aminoglycoside, ticarcillin-clavulanate and quinolone. In the order of frequency administered the oral antibiotics were: second and third-generation cephalosporins, cotrimoxazole, amoxicillin-clavulanate, amoxicillin, first-generation cephalosporins and quinolone.
Within the first 5 days of admission, all patients underwent renal ultrasonography (US). Following completion of antimicrobial therapy and within a month from fever initiation, all patients underwent VCUG to rule out VUR. VUR was classified according to the international grading system.19 The study population included children under the age of 2 years hospitalised from 1 January 2002 to 31 December 2004, prior to the latest American Academy of Pediatrics (2016) and National Institute for Health and Care Excellence (2009) clinical practice guidelines20 21 that repealed universal conduction of VCUG in first episodes of UTI in this age group.
Out of 148 patients, there were 9 (6.1%) with grade 1 or 2 VUR and 21 (14.2%) with grade 3 or 4. Renal ultrasound was abnormal in 43 patients (29.0%) and was indicative of acute pyelonephritis. It is noteworthy that all patients with obstructive uropathy (n=10) were initially excluded, according to the study selection criteria.
All patients underwent DMSA scan 6 months after the first episode of UTI (repeat DMSA). Permanent renal lesions on the follow-up scan were defined as volume loss in conjunction with wedge-shaped defects or thinning and flattening of the cortex.22 DMSA was graded as follows: grade 1: normal (absence of lesions); grade 2: with mild lesions (defect only in one kidney); grade 3: with moderate lesions (extended defect in one kidney or mild lesions in both kidneys); and grade 4: with severe lesions (extended parenchymal lesions in both kidneys). Moderate/severe permanent renal lesions on DMSA scan were defined as ≥3rd grade.23
STATA for Windows V.8.5 (StataCorp, Texas, USA, 2006) was used for data management and statistical analyses. All data were expressed as median (range) in numerical variables and as actual numbers and proportions in categorical variables. Categorical data were compared using Fisher’s exact test. Linear regression analysis was consecutively performed to evaluate the association between the presence of permanent renal lesions on DMSA and FAT. Multiple regression analysis was used to assess and validate the effect of different independent variables on renal lesion formation. P<0.05 was considered significant.
The population characteristics of our patients have been previously published.18 In our study population, the median age (range) was 2.4 months (range: 11 days to 24 months). Of these, 54 (36.5%) were boys and 94 (63.5%) were girls. The median (range) FAT was 24 (1–96) hours.
Thirty-four children (22.97% of the total population) had permanent renal lesions on repeat DMSA scanning 6 months after the first episode of UTI. Of these, 23 (15.5%) had mild lesion (grade 2), 10 (6.7%) had moderate lesion (grade 3) and 1 child (0.6%) had severe (grade 4) lesion. A total of 135 children (91.2%) had fever for <48 hours after the initiation of antibiotic treatment (FAT), while 13 (8.8%) had fever for more than 48 hours. Of the children who had FAT for <48 hours, 107 (79.25%) had normal DMSA 6 months after the UTI, while 28 (20.74%) had permanent renal lesions. Among the latter group, 18 children had mild, 9 had moderate and 1 had severe lesions. Of the 13 children who had FAT>/48 hours, 7 (53.85%) had normal DMSA scan and 6 (46.15%) had permanent lesions (5 had mild and 1 moderate) (table 1A).
Prevalence of permanent renal lesions according to FAT
Table 1A shows the proportion of children with normal repeat DMSA scan and those who had permanent renal lesions according to FAT. As shown in table 1A, the likelihood of permanent renal lesions was significantly increased when FAT was >/48 hours (p=0.048). Moreover, the severity of permanent renal lesions was also significantly increased when FAT was >/48 hours. This was apparent when lesions were classified as mild and moderate/severe (p=0.046) (table 1B).
Characteristics of patients with FAT prolongation
The characteristics of patients with prolonged FAT>/48 hours are subsequently analysed in comparison with the rest of the study group (table 2). Patients with FAT>/48 hours were older in age (p=0.01) and had higher ANC levels (p=0.042). It is noteworthy that prolonged FAT>/48 hours was not associated with the presence of FBT prolongation (>/72 hours), the presence of VUR, US indicative of acute pyelonephritis and the presence of non-E. coli bacteria (table 2).
When studying FAT prolongation (>/48 hours) according to the different types of isolated microorganisms in the urine culture, it was found that in both groups with or without prolonged FAT, E. coli bacteria were equally predominant (84.6% vs 88.1%), while other microorganisms (ie, Proteus, Klebsiella, Enterobacter and Pseudomonas) were in equally small proportions (15.4% vs 11.9%). FAT prolongation was not associated with the type of different isolated microorganisms (Fisher’s exact test, p=0.169).
Factors affecting the severity of permanent renal lesions in repeat DMSA
During univariate linear regression analysis, FAT>/48 hours was significantly associated with the development of renal lesions (R2=0.043, F(1,121)=5.47, p=0.021).
However, when examining the effect of different variables on permanent renal lesion formation during multiple regression models (table 3), among the factors that significantly affected permanent renal lesion development were FBT>/72 hours (p=0.022), patients’ age (p=0.031), CRP (p=0.027) and the presence of VUR (p<0.001), but not FAT (p=0.15). Thus, with the addition of VUR, age, CRP and FBT, the effect of FAT on permanent renal lesion development disappeared (p=0.15).
The present study examines the duration of FAT in childhood febrile UTI and its association with the development of permanent renal lesions in DMSA 6 months after the first UTI episode. The permanent renal lesions, previously called renal scars, were usually mild DMSA uptake defects, possibly lacking prognostic value. However, these lesions might identify those children who are at risk for pyelonephritic damage and subsequent renal scarring,24 whereas children with normal US and DMSA scan after a UTI have a negligible risk of developing a scar after their fourth birthday.25 All repeat DMSA scans were compared with the acute DMSA scans of the patients to assess the development of renal lesions. Although not always possible, the major advantage of an acute DMSA is the identification of primary lesions and their differentiation from secondary lesions.4
It was found that both the likelihood of formation and the severity of these permanent lesions were significantly increased when FAT was >/48 hours. There are very limited previous studies on the role of FAT on the formation of permanent renal lesions after the first episode of childhood febrile UTI, and no previous study, to our knowledge, evaluating FAT in relation to FBT and other factors.
The treatment response time in paediatric febrile UTI is determined by the resolution of fever and generally occurs within the first 24–48 hours of initiation of the appropriate antibiotic treatment.7 16 In the only previously existing study of Hoberman et al,7 children with FAT<36 hours had renal scarring on repeat DMSA rarer than those who responded later (14.0% vs 40.0% children, p=NS), but the difference was not statistically significant. In our population, the cut-off limit of 48 hours yielded a significant relation of permanent kidney damage with prolongation of fever.
Although there are no further previous data on the effect of FAT on repeat DMSA and the formation and severity of permanent renal lesions, there are three studies on acute DMSA findings. According to Fernández-Menéndez et al and Ansari Gilani et al, children with abnormal acute phase DMSA were found to have an average FAT of 29.7 and 47.4 hours, respectively.9 10 Furthermore, when FAT was delayed by >24 hours, renal lesions in the acute DMSA were more common. Since it is accepted that permanent renal lesions after one episode of febrile UTI develop at the sites of acute lesions, the results of these studies might support our finding that an FAT>/48 hours seems to be associated with permanent renal lesion formation.
In contrast with the above studies, in a study by Lee et al, FAT was not related to the formation of renal scars in infants diagnosed with pyelonephritis on acute phase DMSA scanning.26 However, in this study, the maximum FAT was only 60 hours, while in our study median (range) FAT was 24 hours (1–96 hours). Moreover, since this latter study mainly concerns the effect of the VUR, it does not provide enough data on the effect of FAT on the development of renal scars and their severity.
Among the factors associated with the prolongation of treatment response time are the presence of a resistant microorganism, host defence defects, high microorganism virulence or the extent of disease progression before treatment.12–14 It is noteworthy that in all of our patients, antibiotic treatment was chosen according to the isolated microorganism’s sensitivity and none of our patients suffered from immunodeficiency. Regarding high microorganism virulence, in both groups with or without FAT prolongation, non-E. coli bacteria were equally predominant (15.3% vs 11.8%) and FAT prolongation was not associated with the type of isolated microorganisms. In this study, we have also shown that FAT prolongation was associated with increased ANC count (p=0.042), which is indicative of the severity of infection and with older patient’s age (p=0.01).27
Fever is accompanied by elevated levels of pyrogenic and proinflammatory cytokines such as interleukin (IL)-1β, IL-6, IL-8 and others, and is associated with prolongation of the inflammatory process in the renal parenchyma, which is associated with tissue destruction.28–34 Furthermore, we have previously shown that the severity of infection, as expressed by the levels of infection indices (ANC), was associated with the presence and severity of acute renal lesions,18 and also was predictive of the development of permanent renal lesions after a first episode of febrile UTI.15
Regarding the effect of age on the development of permanent renal lesions, in the multiple linear regression analysis of the present study, it was found that increasing age was predictive of the development of renal lesions. Previous studies on this matter had conflicting results. Our findings are in agreement with the studies by Doganis et al,12 Garin et al 35 and Jakobsson et al 5 who showed that the frequency of acute renal lesions during pyelonephritis increased with advancing age. However, Benador et al 36 showed that infants are not more vulnerable to develop permanent renal lesions, but there is overdiagnosis of renal lesions after UTIs due to the presence of congenital renal lesions.
We have previously shown15 that among the factors associated with the development and severity of permanent renal lesions are the delay of treatment initiation (FBT>/72 hours), the presence and severity of vesicoureteral reflux and the severity of infection (as expressed by increased procalcitonin levels and ANC). It seems that dilating VUR remains the strongest risk factor. Also, when the effect of FAT on permanent renal lesion formation was evaluated in relation to other confounders, such as prolonged fever >/72 hours before treatment initiation, the presence of VUR, patients’ age and severity of infection was of less importance.
In the present study, it was shown that prolongation of FAT for >/48 hours was significantly associated with the development of permanent renal lesions in the univariate linear regression. Nevertheless, when validating the effect of FAT>/48 hours on permanent renal lesion development during multiple regression models, we have shown that the effect of FAT unexplainably disappears with the inclusion of other confounders, such as the delay in treatment initiation for >/72 hours, the presence of VUR, CRP levels and older patients’ age. More studies on FAT and other confounders on the development of permanent renal lesions after a first episode of childhood febrile UTI are needed to verify the above findings. However, according to the findings of this study, an FAT>/48 hours and especially an FAT>/72 hours should be considered as a risk factor for permanent renal lesion development. Therefore, parents should be alerted to this possibility and seek prompt medical attention. Early diagnosis and appropriate empirical management of a febrile UTI should be provided to all paediatric patients.
Regarding VUR, in the multiple regression models of this study it was shown that VUR was the most significant factor affecting permanent renal lesion development. A previous study agreed with these findings,37 while another study25 showed that VUR was the only risk factor for renal scar development and the risk increased in parallel with the severity of VUR.27
Strengths and limitations of the study
An important strength of the study was the reliability of the diagnosis of UTI, since urine cultures were obtained by suprapubic aspiration or urinary catheterisation in the vast majority of children (approximately 90%). The main limitations of this study include the small number of patients with severe permanent renal lesions, reducing the statistical significance of our findings, and its retrospective nature. However, our study population was representative of the Greek paediatric population since it was drawn from a Pediatric University Department of a tertiary children’s hospital, which is a referral centre for central and southern Greece. Another limitation is that acquired permanent renal lesions sometimes are difficult to distinguish from congenital. The primary renal damage in these patients is usually associated with higher grades of VUR,23 and both findings may be the result of embryonal abnormality of the ureteral bud. Children with congenital anomalies of the kidney and urinary tract are mainly boys with global parenchymal reduction, that is, small kidneys without focal scars.24
In the present study, duration of fever≥48 hours after initiation of treatment of a first episode of febrile UTI in children younger than 2 years old was a significant predictive factor of the development and severity of permanent renal lesions. However, other parameters, such as prolonged duration of FAT (>/72 hours), the presence of vesicoureteral reflux, the severity of infection and older patients’ age, seem to affect the development of permanent renal lesions more significantly.
Contributors All authors have contributed equally to this manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None decalred.
Patient consent Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.