- Email: [email protected]
Relationship Among Vesicoureteral Reflux, Urinary Tract Infection and Renal Damage in Children
Relationship Among Vesicoureteral Reflux, Urinary Tract Infection and Renal Damage in Children Svante Swerkersson,* Ulf Jodal, Rune Sixt, Eira Stokland and Sverker Hansson From the Departments of Pediatrics, Pediatric Clinical Physiology and Pediatric Radiology, Pediatric Uro-Nephrologic Center, Göteborg University, Göteborg, Sweden
Purpose: We studied the relationship among vesicoureteral reflux, urinary tract infection and permanent renal damage in children. Materials and Methods: We retrospectively analyzed 303 children younger than 2 years with a first time, nonobstructive, culture verified urinary tract infection. The protocol included ultrasonography and voiding cystourethrography within 3 months after urinary tract infection, and 99mtechnetium dimercapto-succinic acid scintigraphy after 1 to 2 years. Results: Vesicoureteral reflux was found in 36 of 163 boys (22%) and in 44 of 140 girls (31%). Of the 303 patients 80 (26%) had permanent renal damage according to dimercapto-succinic acid scintigraphy. The rate of abnormality increased significantly with grade of vesicoureteral reflux in boys and girls. The relative risk of renal damage was significantly increased in patients with vesicoureteral reflux grade II and higher. Maximum C-reactive protein concentration, maximum temperature during urinary tract infection, presence of vesicoureteral reflux and febrile recurrences of urinary tract infection were significantly related to permanent renal damage. In stepwise logistic regression vesicoureteral reflux was the only independent variable for boys, while C-reactive protein and vesicoureteral reflux were independent factors for girls. Conclusions: There was a significant relationship between grade II vesicoureteral reflux and higher and permanent renal damage in boys and girls. However, while the association between renal damage and vesicoureteral reflux was evident in boys, the role of urinary tract infection and renal inflammation seemed to be equally or more important in girls. These findings support the concept that renal damage is associated with vesicoureteral reflux and is often congenital in boys, while in girls it is more related to urinary tract infection with vesicoureteral reflux as a reinforcing factor. Key Words: urinary tract infections, vesico-ureteral reflux, C-reactive protein, technetium Tc 99m dimercaptosuccinic acid, urodynamics
study the same patients were used to determine the relationship of VUR and UTI to permanent renal damage, and to elucidate the possible differences between boys and girls.
rinary tract infection is common in children. In a Swedish national study of children younger than 2 years the estimated minimum incidence of UTI was 2.2% for boys and 2.1% for girls.1 Most UTIs in infants are febrile, indicating renal involvement. Renal infection may lead to permanent damage.2,3 In a study of 157 children younger than 6 years with a first time UTI 59 (38%) had damage on 99mtechnetium DMSA scintigraphy at 1 to 2-year followup.4 The role of VUR in the pathogenesis of permanent renal damage in children with UTI has been debated in recent years. A different pattern between genders has been proposed, with congenital scarring with dysplastic features associated with prenatal VUR in boys and acquired scarring secondary to UTI with or without VUR in girls.5– 8 In a previous study we analyzed the ability of early DMSA scanning to predict the presence of VUR.9 We concluded that DMSA scintigraphy in infants with UTI may replace voiding cystourethrography as a first-line investigation. In this
U
MATERIALS AND METHODS The study population consisted of 303 children younger than 2 years presenting emergently with first time, nonobstructive, symptomatic UTI. Subjects were selected by a search of the files of the UTI clinic for patients who within 3 months of the UTI had been investigated by ultrasonography, VCU and DMSA scan, with a second DMSA scan at 1 to 2 years. Children with known urogenital or anorectal malformations or neurological disease were excluded, as were those with suspected obstruction on ultrasonography. Of the patients 163 were boys, all uncircumcised, with a median age of 3.1 months (range 5 days to 19.9 months) and 140 were girls with a median age of 8.5 months (5 days to 22.6 months). Maximum temperature and maximum CRP concentration in serum at the first UTI were documented. Bacteriuria was defined as any growth of bacteria in urine from suprapubic bladder aspiration, or greater than 100,000 colony forming units of a single type in urine from a bag, midstream or catheter sample. Escherichia coli was isolated in 277 of the children, Klebsiella/Enterobacter in 13, enterococci in 7, and Staphylococcus aureus, coagulase negative staphylo-
Submitted for publication November 22, 2006. Supported by grants from the Frimurare-Barnhusdirektionen, Swedish Research Council, and Health and Medical Care Executive Board of the Västra Götaland Region. * Correspondence: The Queen Silvia Children’s Hospital, SE-416 85 Goteborg, Sweden (telephone: 46-31-3434000; FAX: 46-31843653; e-mail: [email protected]).
0022-5347/07/1782-0647/0 THE JOURNAL OF UROLOGY® Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION
647
Vol. 178, 647-651, August 2007 Printed in U.S.A. DOI:10.1016/j.juro.2007.04.004
648
VESICOURETERAL REFLUX, INFECTION AND RENAL DAMAGE
cocci, Pseudomonas, Serratia and Citrobacter in 1 each. Bacterial typing was not performed in 1 girl with a positive nitrite test and significant bacterial growth on dipslide. Static renal scintigraphy was performed 3 to 4 hours after injection of DMSA at a dose of 1 MBq/kg body weight (minimum 15 MBq).5 Images were obtained in 3 projections, ie 1 posterior and 2 oblique, with 300,000 counts in the posterior view. A focal reduction or absence of uptake in 1 or more areas in the kidney was considered abnormal and indicative of renal damage. A kidney with a relative function of 44% or less was also classified as abnormal. The radiological examinations were performed according to standard procedures of the Pediatric Radiology Department. Reflux was graded I to V according to the recommendations of the International Reflux Study in Children.10 Children with dilating VUR (grades III to V) were given antibacterial prophylaxis consisting of a single dose of 0.5 to 1 mg/kg body weight trimethoprim daily, and were examined by a second VCU at 1 to 2 years. Analysis of the DMSA scan was performed by a clinical physiologist (RS), and VCU was performed by a radiologist (ES), always without knowledge of the other data. For statistical comparisons between groups we used Wilcoxon’s 2-sample test. The Mantel-Haenszel chi-square test was used to analyze the trend in a contingency table. Spearman’s rank correlation coefficient was used for correlational analyses. Relative risks with 95% confidence intervals were calculated to detect differences between VUR grades. Stepwise logistic regression was used for multivariate purposes. In cases of bilateral VUR the most severe grade was used. RESULTS Character of UTI Information about temperature at first UTI was lacking in 2 patients, and data about CRP were lacking in 1. Of 161 boys 118 (73%) had a febrile UTI (temperature 38.5C or greater), as did 128 of 140 girls (91%). A total of 232 children (77%) had a maximum CRP of 20 mg/l or greater. Median CRP was 49 mg/l (range 0 to 290) in boys and 65 mg/l (0 to 290) in girls, a difference that was significant (p ⬍0.05). Despite a maximum temperature of less than 38.5C, 21 children had a CRP of 20 mg/l or greater (median 60, range 20 to 148). Of these 21 patients 14 were younger than 1 month, and all were male. Vesicoureteral Reflux Reflux was found in 22% of the boys (36 of 163) and in 31% of the girls (44 of 140, table 1). Dilating VUR (grades III to V) was found significantly more often in boys (22 of 36 with VUR) than in girls (14 of 44 with VUR, p ⬍0.01). Only 1 boy had grade V VUR.
FIG. 1. Level of C-reactive protein (mg/l) in relation to VUR grade. Results are expressed as box plots indicating medians with lower and upper quartiles. Whiskers show 10th and 90th percentiles.
There was a significant relationship between maximum CRP at first UTI and grade of VUR in boys (p ⬍0.05) and girls (p ⬍0.01, fig. 1). Also, UTI recurrence with fever (38.5C or greater) occurred in 36 children (12%) during followup out to the second DMSA scan at 1 to 2 years (table 2). The risk for new febrile UTIs increased with the presence and severity of VUR (p ⬍0.001). Followup DMSA Scintigraphy At the followup examination 80 of 303 patients (26%) had abnormal DMSA scintigraphy. The rate of abnormality was 19% (43 of 223 patients) in those without demonstrable VUR. There was a significant relationship between DMSA abnormality and the presence and severity of VUR (p ⬍0.001, table 1). The relative risk of renal abnormality with 95% confidence limits in relation to VUR grade is shown in figure 2 (grade I, 1.20 [0.43 to 3.35]; grade II, 2.17 [1.33 to 3.56]; grade III, 2.50 [1.55 to 4.01] and grades IV to V, 4.61 [3.23 to 6.57]). There was a significantly increased risk in males and females with VUR grade II and higher. The maximum temperature at first UTI correlated with renal abnormalities on followup DMSA scan (p ⬍0.05). There was also a significant relationship between the maximum CRP at first UTI and renal abnormalities on followup DMSA scan (p ⬍0.001, table 3). Children with recurrent UTI had permanent renal damage significantly more often than those without recurrent UTI (p ⬍0.01, table 4). Age and gender were not significantly related to the presence of renal damage. There were 15 children (9 boys and 6 girls) with dilating VUR without renal damage on followup DMSA scintigraphy.
TABLE 2. Febrile UTI recurrences TABLE 1. Abnormal DMSA scintigraphy at followup according to VUR grade at first UTI VUR Grade
No. Boys
No. Girls
No. Abnormal DMSA Scintigrams (%)
No VUR I II III IV–V
127 3 11 16 6
96 10 20 11 3
43 (19) 3 (23) 13 (42) 13 (48) 8 (89)
No. Recurrences 0 1 2 3 Totals
No. VUR Grade No VUR
I
II
III
IV–V
202 20 1 0
10 1 2 0
27 3 1 0
21 3 1 2
7 1 0 1
223
13
31
27
9
Children with grades III to V VUR were given antibacterial prophylaxis.
VESICOURETERAL REFLUX, INFECTION AND RENAL DAMAGE
649
TABLE 4. Abnormal DMSA scintigraphy at followup based on number of febrile UTI recurrences
FIG. 2. Relative risk and 95% CI of abnormality on followup DMSA scintigraphy in infants with different grades of VUR (I to V) compared to infants without demonstrable VUR.
Of these children 14 had VUR grade III, and 1 girl had bilateral VUR grade IV. In 9 of these children the initial (early) DMSA scan was abnormal. All patients except 1 had a high fever at first UTI and all except 1 had UTI caused by E. coli. Two of these 15 children had febrile recurrences during followup before the second DMSA scan was performed. A new discrete focal scar was detected in a previously unscarred kidney in 1 of these patients, and there was progression of a preexisting scar in 1. Relationship of UTI, VUR and Permanent Renal Scarring Reflux grade, temperature, CRP level during first UTI and recurrence of febrile UTI were all significantly related to permanent renal damage at DMSA scintigraphy at 1 to 2 years. These variables were selected as possible independent predictors in a stepwise logistic regression model, and VUR, CRP and recurrent UTI entered the model as independent variables. When boys were analyzed separately VUR was the only independent variable (p ⬍0.0001), and when only girls were analyzed CRP levels at first UTI (p ⬍0.001) and VUR (p ⬍0.05) were significant independent factors. In girls without VUR CRP was the only independent variable, and in boys without VUR no independent variable was found. DISCUSSION Most patients in this study (82%) had febrile UTI. These patients constitute a representative group of children up to 2 years old with nonobstructive UTI. Among the patients 90% of the boys and 70% of the girls were younger than 1 year. The dominance of males during the first 6 months of life is characteristic of a population where circumcision is not a common procedure and is in accordance with previous studies.1,11 Reflux was present in 22% of the boys and in 31% of the girls, a finding that is also in agreement with previous studies.12 Finally, boys had higher grades of VUR than girls.
TABLE 3. Abnormal DMSA scintigraphy at followup based on maximum CRP at first UTI CRP (mg/l)
No. Boys (%)
No. Girls (%)
Less than 20 20–49 50–99 100 or Greater
7 (15) 9 (26) 10 (22) 12 (34)
3 (13) 4 (13) 16 (38) 19 (44)
Totals (av)
38 (23)
42 (30)
CRP analysis was not available in 1 boy.
No. Recurrences
No. Boys (%)
No. Girls (%)
0 1 2 or More
32 (22) 4 (33) 2 (67)
32 (26) 6 (38) 4 (80)
Totals (av)
38 (23)
42 (30)
The finding that 19% of the infants without VUR had abnormal DMSA scans confirms previous reports that VUR is not a prerequisite for renal damage.9,13 Furthermore, the relative risk for infants with grade I VUR was the same as for infants without VUR. In patients with grade II VUR the relative risk was significantly higher (2.17, 95% confidence limits 1.33 to 3.56). At the other end of the spectrum 8 of 9 infants with grades IV to V VUR had renal abnormality. It is interesting that 14 of 27 patients with grade III and 1 of 8 with grade IV reflux had normal kidneys on followup DMSA scan. The impact of UTI on renal outcome was also evident in this study. However, this effect was only observed for girls. Maximum CRP at first UTI and febrile reinfections were related to an abnormal DMSA scan on followup. The importance of recurrent infections has been observed in previous studies. In a series of 74 children with renal scarring diagnosed by urography girls with acquired scarring had significantly more febrile recurrences than a matched group of girls without scarring.8 In another study of 111 women with a history of childhood UTI the development of scarring and the progression of established scars were related to the number of febrile recurrences.14 Thus, prompt treatment of UTI and prevention of new infections to avoid development or progression of renal damage are important. We can speculate that this approach is essential for both genders, although the smaller risk of recurrent infections in boys makes it difficult to prove for boys. We also found a relationship between the maximum CRP level at first UTI and the presence and grade of VUR. This finding is important, since it suggests that VUR in itself reinforces the inflammatory response, and thereby has the potential to increase the risk of permanent kidney damage following UTI. Antibiotic prophylaxis has traditionally been used to prevent new infections in patients at risk. Despite our prescription of prophylactic drugs to children with dilating VUR (grades III to V), this group had significantly more febrile recurrences than those with low grade or no VUR. There is also a lack of controlled studies showing that prophylaxis is superior to prompt treatment of recurrent UTIs as far as renal outcome is concerned.15 Finally, there is a need for more information about which patients really benefit from long-term prophylaxis as well as the optimal duration of treatment. Whether active treatment of VUR hinders or inhibits development or progress of scarring has been debated for several years.16 In a meta-analysis of children with UTI VUR was a weak predictor of renal damage, although it was not possible to extract sufficient data to evaluate whether more severe grades of VUR were better predictors of renal damage compared to minor grades.17 In a 10-year followup study of patients with grades III and IV VUR randomized to
650
VESICOURETERAL REFLUX, INFECTION AND RENAL DAMAGE
either surgical or medical treatment there was no difference with regard to identification of new renal scars.18 New scars developed mostly in children younger than 5 years, and more frequently in those with grade IV than with grade III reflux. In a meta-analysis of randomized controlled trials comparing antibiotics and surgery for VUR it was not clear whether any intervention was beneficial.19 During recent years an endoscopic technique to treat VUR has been introduced.20 This procedure is less invasive and has replaced reimplantation at many centers. However, a major problem is that no known controlled study has been published evaluating the effect of this procedure on the frequency of febrile recurrences, or on the long-term outcome of kidney development or function. Thus, the indications for active treatment vs antibiotic prophylaxis need to be addressed in controlled randomized trials, which also should include a group managed by observation and prompt treatment of each UTI as it occurs. The Swedish Reflux Study is an ongoing project comparing these 3 alternatives, with febrile reinfections, renal damage on DMSA scan and VUR outcome as end points. CONCLUSIONS This study reveals an association between VUR grade and permanent renal damage in boys and girls. In girls the grade of inflammation at first UTI was also a significant factor, and one that might be even more important than reflux grade. The study lends support to the concept that renal damage is often congenital and is associated with VUR in boys, whereas in girls it is more often acquired and related to the severity of inflammatory reaction.
6. 7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
ACKNOWLEDGMENTS 17.
Gunnar Ekeroth and Emma Jaensson assisted with the statistical analysis.
Abbreviations and Acronyms CRP DMSA UTI VCU VUR
⫽ ⫽ ⫽ ⫽ ⫽
C-reactive protein dimercapto-succinic acid urinary tract infection voiding cystourethrography vesicoureteral reflux
REFERENCES 1.
Jakobsson B, Esbjorner E and Hansson S: Minimum incidence and diagnostic rate of first urinary tract infection. Pediatrics 1999; 104: 222. 2. Glauser MP, Meylan P and Bille J: The inflammatory response and tissue damage. The example of renal scars following acute renal infection. Pediatr Nephrol 1987; 1: 615. 3. Svanborg C, Bergsten G, Fischer H, Frendeus B, Godaly G, Gustafsson E et al: The ‘innate’ host response protects and damages the infected urinary tract. Ann Med 2001; 33: 563. 4. Stokland E, Hellstrom M, Jacobsson B, Jodal U and Sixt R: Renal damage one year after first urinary tract infection: role of dimercaptosuccinic acid scintigraphy. J Pediatr 1996; 129: 815. 5. Risdon RA: The small scarred kidney of childhood. A congenital or an acquired lesion? Pediatr Nephrol 1987; 1: 632.
18.
19.
20.
Risdon RA: The small scarred kidney in childhood. Pediatr Nephrol 1993; 7: 361. Yeung CK, Godley ML, Dhillon HK, Gordon I, Duffy PG and Ransley PG: The characteristics of primary vesico-ureteric reflux in male and female infants with pre-natal hydronephrosis. Br J Urol 1997; 80: 319. Wennerstrom M, Hansson S, Jodal U and Stokland E: Primary and acquired renal scarring in boys and girls with urinary tract infection. J Pediatr 2000; 136: 30. Hansson S, Dhamey M, Sigstrom O, Sixt R, Stokland E, Wennerstrom M et al: Dimercapto-succinic acid scintigraphy instead of voiding cystourethrography for infants with urinary tract infection. J Urol 2004; 172: 1071. Lebowitz RL, Olbing H, Parkkulainen KV, Smellie JM and Tamminen-Mobius TE: International system of radiographic grading of vesicoureteric reflux. International Reflux Study in Children. Pediatr Radiol 1985; 15: 105. Hansson S, Bollgren I, Esbjorner E, Jakobsson B and Marild S: Urinary tract infections in children below two years of age: a quality assurance project in Sweden. The Swedish Pediatric Nephrology Association. Acta Paediatr 1999; 88: 270. Jacobson SH, Hansson S and Jakobsson B: Vesico-ureteric reflux: occurrence and long-term risks. Acta Paediatr, suppl., 1999; 88: 22. Rushton HG: The evaluation of acute pyelonephritis and renal scarring with technetium 99m-dimercaptosuccinic acid renal scintigraphy: evolving concepts and future directions. Pediatr Nephrol 1997; 11: 108. Martinell J, Lidin-Janson G, Jagenburg R, Sivertsson R, Claesson I and Jodal U: Girls prone to urinary infections followed into adulthood. Indices of renal disease. Pediatr Nephrol 1996; 10: 139. Williams G, Lee A and Craig J: Antibiotics for the prevention of urinary tract infection in children: a systematic review of randomized controlled trials. J Pediatr 2001; 138: 868. Jodal U, Hansson S and Hjalmas K: Medical or surgical management for children with vesico-ureteric reflux? Acta Paediatr, suppl., 1999; 88: 53. Gordon I, Barkovics M, Pindoria S, Cole T and Woolf AS: Primary vesicoureteric reflux as a predictor of renal damage in children hospitalized with urinary tract infection: a systematic review and meta-analysis. J Am Soc Nephrol 2003; 14: 739. Olbing H, Smellie JM, Jodal U and Lax H: New renal scars in children with severe VUR: a 10-year study of randomized treatment. Pediatr Nephrol 2003; 18: 1128. Wheeler D, Vimalachandra D, Hodson EM, Roy LP, Smith G and Craig JC: Antibiotics and surgery for vesicoureteric reflux: a meta-analysis of randomised controlled trials. Arch Dis Child 2003; 88: 688. Lackgren G, Wahlin N and Stenberg A: Endoscopic treatment of children with vesico-ureteric reflux. Acta Paediatr, suppl., 1999; 88: 62.
EDITORIAL COMMENTS The specific aim of the authors was to study the relationship between urinary tract infection, vesicoureteral reflux and acquired renal cortical scarring in a retrospective analysis of 303 children younger than 2 years with a first time, culture documented, symptomatic urinary tract infection. Each patient underwent DMSA renal scintigraphy within 3 months of presenting for UTI, as well as after 1 to 2 years of rigorous followup while on antibiotic prophylaxis. The patients studied represent the same cohort that formed the basis of an earlier series supporting the use of DMSA renal scanning as the first imaging tool in the evaluation of UTI because of its ability to indicate the presence of vesicoureteral reflux based
VESICOURETERAL REFLUX, INFECTION AND RENAL DAMAGE on abnormalities seen in the renal parenchyma consistent with acute pyelonephritis. In that study 156 of the 303 children (51%) presenting with a culture proved UTI and fever had an abnormal DMSA scan initially, thus, excluding the remaining children from being considered as having radiologically proved acute pyelonephritis. The authors found a significant relationship between the presence of acute lesions and the presence and severity of VUR (p ⬍0.001), and in particular there was an increased risk of DMSA scan abnormalities in boys and girls with grade III and higher VUR (reference 9 in article). In the current study 26% of children presenting with symptomatic UTI had DMSA renal scan abnormalities on followup 2 years later. It is noteworthy that 49% of the children were unlikely to form scars simply on the basis of negative DMSA renal scan abnormalities on initial presentation. Taking this finding into account, it can be estimated that the incidence of renal scarring in children with initially positive renal scans is 80 of 156, or 51%, which is much higher than what has been reported in other studies. Thus, this patient population is heterogeneous (includes patients without radiographically proved acute pyelonephritis) and may include a cohort of children with a greater predisposition to renal scarring following pyelonephritis than is typically seen. The authors found that DMSA renal scan abnormalities were correlated with the presence of VUR grade II and higher in children of both genders, with maximum recorded body temperature and maximum C-reactive protein level at first UTI, and with the number of recurrent UTIs while on prophylaxis. However, in a stepwise logistic regression analysis searching for independent variables only VUR was strongly correlated with renal scarring, with the strongest correlation occurring in boys. The relative risk that VUR was associated with renal scan abnormalities increased significantly by grade, from 1.20 for grade I to 4.61 for grades IV and V. It is noteworthy that boys were more likely to have higher grades of reflux, a finding that raises suspicion that some of the abnormalities noted at 2 years might in fact have been present at birth. An additional interesting observation can be made in the cohort of children (15 of 80) who had dilating vesicoureteral reflux (grades III to IV) and initially positive DMSA scans
651
(including 2 children who also had recurrent UTI), who were obviously at low risk for renal scarring on the basis of normal followup DMSA renal scans. Clearly, some children are inherently protected from the ill effects of acute pyelonephritis by mechanisms that are as yet unclear to us. Hans G. Pohl Department of Pediatric Urology George Washington University Children’s National Medical Center Washington, D. C.
The authors seem to have shown that there is an association between reflux and recurrent urinary tract infection, and reflux and abnormal nuclear renal scans. Of the children 12% had a febrile UTI recurrence, and this outcome correlated with the presence and severity of reflux. Abnormalities in the renal scan occurred more often in patients who had reflux. However, the question remains as to whether these findings are due to cause and effect, or are merely associations. Reading this article, one is interested in the CRP data. In the population studied CRP levels correlated with DMSA abnormalities. Analyzing this data, it would seem that girls with a CRP of less than 50 and boys with a CRP of less than 20 should not receive a DMSA scan, since the chance of an abnormality on this study at those CRP levels is quite low. The data presented in figure 1 led to another question. If these data were presented in a scatter plot, would there be a CRP level low enough that could confidently correlate with absolutely no cases of reflux? If so, many children might be spared the unpopular voiding cystourethrogram. All in all, I believe that the authors do a service to our specialty and to pediatricians at large because their findings countermand some of the growing voices that suggest urinary reflux is not a significant condition. Richard N. Schlussel Department of Urology Columbia University Medical Center New York, New York
Purpose: We studied the relationship among vesicoureteral reflux, urinary tract infection and permanent renal damage in children. Materials and Methods: We retrospectively analyzed 303 children younger than 2 years with a first time, nonobstructive, culture verified urinary tract infection. The protocol included ultrasonography and voiding cystourethrography within 3 months after urinary tract infection, and 99mtechnetium dimercapto-succinic acid scintigraphy after 1 to 2 years. Results: Vesicoureteral reflux was found in 36 of 163 boys (22%) and in 44 of 140 girls (31%). Of the 303 patients 80 (26%) had permanent renal damage according to dimercapto-succinic acid scintigraphy. The rate of abnormality increased significantly with grade of vesicoureteral reflux in boys and girls. The relative risk of renal damage was significantly increased in patients with vesicoureteral reflux grade II and higher. Maximum C-reactive protein concentration, maximum temperature during urinary tract infection, presence of vesicoureteral reflux and febrile recurrences of urinary tract infection were significantly related to permanent renal damage. In stepwise logistic regression vesicoureteral reflux was the only independent variable for boys, while C-reactive protein and vesicoureteral reflux were independent factors for girls. Conclusions: There was a significant relationship between grade II vesicoureteral reflux and higher and permanent renal damage in boys and girls. However, while the association between renal damage and vesicoureteral reflux was evident in boys, the role of urinary tract infection and renal inflammation seemed to be equally or more important in girls. These findings support the concept that renal damage is associated with vesicoureteral reflux and is often congenital in boys, while in girls it is more related to urinary tract infection with vesicoureteral reflux as a reinforcing factor. Key Words: urinary tract infections, vesico-ureteral reflux, C-reactive protein, technetium Tc 99m dimercaptosuccinic acid, urodynamics
study the same patients were used to determine the relationship of VUR and UTI to permanent renal damage, and to elucidate the possible differences between boys and girls.
rinary tract infection is common in children. In a Swedish national study of children younger than 2 years the estimated minimum incidence of UTI was 2.2% for boys and 2.1% for girls.1 Most UTIs in infants are febrile, indicating renal involvement. Renal infection may lead to permanent damage.2,3 In a study of 157 children younger than 6 years with a first time UTI 59 (38%) had damage on 99mtechnetium DMSA scintigraphy at 1 to 2-year followup.4 The role of VUR in the pathogenesis of permanent renal damage in children with UTI has been debated in recent years. A different pattern between genders has been proposed, with congenital scarring with dysplastic features associated with prenatal VUR in boys and acquired scarring secondary to UTI with or without VUR in girls.5– 8 In a previous study we analyzed the ability of early DMSA scanning to predict the presence of VUR.9 We concluded that DMSA scintigraphy in infants with UTI may replace voiding cystourethrography as a first-line investigation. In this
U
MATERIALS AND METHODS The study population consisted of 303 children younger than 2 years presenting emergently with first time, nonobstructive, symptomatic UTI. Subjects were selected by a search of the files of the UTI clinic for patients who within 3 months of the UTI had been investigated by ultrasonography, VCU and DMSA scan, with a second DMSA scan at 1 to 2 years. Children with known urogenital or anorectal malformations or neurological disease were excluded, as were those with suspected obstruction on ultrasonography. Of the patients 163 were boys, all uncircumcised, with a median age of 3.1 months (range 5 days to 19.9 months) and 140 were girls with a median age of 8.5 months (5 days to 22.6 months). Maximum temperature and maximum CRP concentration in serum at the first UTI were documented. Bacteriuria was defined as any growth of bacteria in urine from suprapubic bladder aspiration, or greater than 100,000 colony forming units of a single type in urine from a bag, midstream or catheter sample. Escherichia coli was isolated in 277 of the children, Klebsiella/Enterobacter in 13, enterococci in 7, and Staphylococcus aureus, coagulase negative staphylo-
Submitted for publication November 22, 2006. Supported by grants from the Frimurare-Barnhusdirektionen, Swedish Research Council, and Health and Medical Care Executive Board of the Västra Götaland Region. * Correspondence: The Queen Silvia Children’s Hospital, SE-416 85 Goteborg, Sweden (telephone: 46-31-3434000; FAX: 46-31843653; e-mail: [email protected]).
0022-5347/07/1782-0647/0 THE JOURNAL OF UROLOGY® Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION
647
Vol. 178, 647-651, August 2007 Printed in U.S.A. DOI:10.1016/j.juro.2007.04.004
648
VESICOURETERAL REFLUX, INFECTION AND RENAL DAMAGE
cocci, Pseudomonas, Serratia and Citrobacter in 1 each. Bacterial typing was not performed in 1 girl with a positive nitrite test and significant bacterial growth on dipslide. Static renal scintigraphy was performed 3 to 4 hours after injection of DMSA at a dose of 1 MBq/kg body weight (minimum 15 MBq).5 Images were obtained in 3 projections, ie 1 posterior and 2 oblique, with 300,000 counts in the posterior view. A focal reduction or absence of uptake in 1 or more areas in the kidney was considered abnormal and indicative of renal damage. A kidney with a relative function of 44% or less was also classified as abnormal. The radiological examinations were performed according to standard procedures of the Pediatric Radiology Department. Reflux was graded I to V according to the recommendations of the International Reflux Study in Children.10 Children with dilating VUR (grades III to V) were given antibacterial prophylaxis consisting of a single dose of 0.5 to 1 mg/kg body weight trimethoprim daily, and were examined by a second VCU at 1 to 2 years. Analysis of the DMSA scan was performed by a clinical physiologist (RS), and VCU was performed by a radiologist (ES), always without knowledge of the other data. For statistical comparisons between groups we used Wilcoxon’s 2-sample test. The Mantel-Haenszel chi-square test was used to analyze the trend in a contingency table. Spearman’s rank correlation coefficient was used for correlational analyses. Relative risks with 95% confidence intervals were calculated to detect differences between VUR grades. Stepwise logistic regression was used for multivariate purposes. In cases of bilateral VUR the most severe grade was used. RESULTS Character of UTI Information about temperature at first UTI was lacking in 2 patients, and data about CRP were lacking in 1. Of 161 boys 118 (73%) had a febrile UTI (temperature 38.5C or greater), as did 128 of 140 girls (91%). A total of 232 children (77%) had a maximum CRP of 20 mg/l or greater. Median CRP was 49 mg/l (range 0 to 290) in boys and 65 mg/l (0 to 290) in girls, a difference that was significant (p ⬍0.05). Despite a maximum temperature of less than 38.5C, 21 children had a CRP of 20 mg/l or greater (median 60, range 20 to 148). Of these 21 patients 14 were younger than 1 month, and all were male. Vesicoureteral Reflux Reflux was found in 22% of the boys (36 of 163) and in 31% of the girls (44 of 140, table 1). Dilating VUR (grades III to V) was found significantly more often in boys (22 of 36 with VUR) than in girls (14 of 44 with VUR, p ⬍0.01). Only 1 boy had grade V VUR.
FIG. 1. Level of C-reactive protein (mg/l) in relation to VUR grade. Results are expressed as box plots indicating medians with lower and upper quartiles. Whiskers show 10th and 90th percentiles.
There was a significant relationship between maximum CRP at first UTI and grade of VUR in boys (p ⬍0.05) and girls (p ⬍0.01, fig. 1). Also, UTI recurrence with fever (38.5C or greater) occurred in 36 children (12%) during followup out to the second DMSA scan at 1 to 2 years (table 2). The risk for new febrile UTIs increased with the presence and severity of VUR (p ⬍0.001). Followup DMSA Scintigraphy At the followup examination 80 of 303 patients (26%) had abnormal DMSA scintigraphy. The rate of abnormality was 19% (43 of 223 patients) in those without demonstrable VUR. There was a significant relationship between DMSA abnormality and the presence and severity of VUR (p ⬍0.001, table 1). The relative risk of renal abnormality with 95% confidence limits in relation to VUR grade is shown in figure 2 (grade I, 1.20 [0.43 to 3.35]; grade II, 2.17 [1.33 to 3.56]; grade III, 2.50 [1.55 to 4.01] and grades IV to V, 4.61 [3.23 to 6.57]). There was a significantly increased risk in males and females with VUR grade II and higher. The maximum temperature at first UTI correlated with renal abnormalities on followup DMSA scan (p ⬍0.05). There was also a significant relationship between the maximum CRP at first UTI and renal abnormalities on followup DMSA scan (p ⬍0.001, table 3). Children with recurrent UTI had permanent renal damage significantly more often than those without recurrent UTI (p ⬍0.01, table 4). Age and gender were not significantly related to the presence of renal damage. There were 15 children (9 boys and 6 girls) with dilating VUR without renal damage on followup DMSA scintigraphy.
TABLE 2. Febrile UTI recurrences TABLE 1. Abnormal DMSA scintigraphy at followup according to VUR grade at first UTI VUR Grade
No. Boys
No. Girls
No. Abnormal DMSA Scintigrams (%)
No VUR I II III IV–V
127 3 11 16 6
96 10 20 11 3
43 (19) 3 (23) 13 (42) 13 (48) 8 (89)
No. Recurrences 0 1 2 3 Totals
No. VUR Grade No VUR
I
II
III
IV–V
202 20 1 0
10 1 2 0
27 3 1 0
21 3 1 2
7 1 0 1
223
13
31
27
9
Children with grades III to V VUR were given antibacterial prophylaxis.
VESICOURETERAL REFLUX, INFECTION AND RENAL DAMAGE
649
TABLE 4. Abnormal DMSA scintigraphy at followup based on number of febrile UTI recurrences
FIG. 2. Relative risk and 95% CI of abnormality on followup DMSA scintigraphy in infants with different grades of VUR (I to V) compared to infants without demonstrable VUR.
Of these children 14 had VUR grade III, and 1 girl had bilateral VUR grade IV. In 9 of these children the initial (early) DMSA scan was abnormal. All patients except 1 had a high fever at first UTI and all except 1 had UTI caused by E. coli. Two of these 15 children had febrile recurrences during followup before the second DMSA scan was performed. A new discrete focal scar was detected in a previously unscarred kidney in 1 of these patients, and there was progression of a preexisting scar in 1. Relationship of UTI, VUR and Permanent Renal Scarring Reflux grade, temperature, CRP level during first UTI and recurrence of febrile UTI were all significantly related to permanent renal damage at DMSA scintigraphy at 1 to 2 years. These variables were selected as possible independent predictors in a stepwise logistic regression model, and VUR, CRP and recurrent UTI entered the model as independent variables. When boys were analyzed separately VUR was the only independent variable (p ⬍0.0001), and when only girls were analyzed CRP levels at first UTI (p ⬍0.001) and VUR (p ⬍0.05) were significant independent factors. In girls without VUR CRP was the only independent variable, and in boys without VUR no independent variable was found. DISCUSSION Most patients in this study (82%) had febrile UTI. These patients constitute a representative group of children up to 2 years old with nonobstructive UTI. Among the patients 90% of the boys and 70% of the girls were younger than 1 year. The dominance of males during the first 6 months of life is characteristic of a population where circumcision is not a common procedure and is in accordance with previous studies.1,11 Reflux was present in 22% of the boys and in 31% of the girls, a finding that is also in agreement with previous studies.12 Finally, boys had higher grades of VUR than girls.
TABLE 3. Abnormal DMSA scintigraphy at followup based on maximum CRP at first UTI CRP (mg/l)
No. Boys (%)
No. Girls (%)
Less than 20 20–49 50–99 100 or Greater
7 (15) 9 (26) 10 (22) 12 (34)
3 (13) 4 (13) 16 (38) 19 (44)
Totals (av)
38 (23)
42 (30)
CRP analysis was not available in 1 boy.
No. Recurrences
No. Boys (%)
No. Girls (%)
0 1 2 or More
32 (22) 4 (33) 2 (67)
32 (26) 6 (38) 4 (80)
Totals (av)
38 (23)
42 (30)
The finding that 19% of the infants without VUR had abnormal DMSA scans confirms previous reports that VUR is not a prerequisite for renal damage.9,13 Furthermore, the relative risk for infants with grade I VUR was the same as for infants without VUR. In patients with grade II VUR the relative risk was significantly higher (2.17, 95% confidence limits 1.33 to 3.56). At the other end of the spectrum 8 of 9 infants with grades IV to V VUR had renal abnormality. It is interesting that 14 of 27 patients with grade III and 1 of 8 with grade IV reflux had normal kidneys on followup DMSA scan. The impact of UTI on renal outcome was also evident in this study. However, this effect was only observed for girls. Maximum CRP at first UTI and febrile reinfections were related to an abnormal DMSA scan on followup. The importance of recurrent infections has been observed in previous studies. In a series of 74 children with renal scarring diagnosed by urography girls with acquired scarring had significantly more febrile recurrences than a matched group of girls without scarring.8 In another study of 111 women with a history of childhood UTI the development of scarring and the progression of established scars were related to the number of febrile recurrences.14 Thus, prompt treatment of UTI and prevention of new infections to avoid development or progression of renal damage are important. We can speculate that this approach is essential for both genders, although the smaller risk of recurrent infections in boys makes it difficult to prove for boys. We also found a relationship between the maximum CRP level at first UTI and the presence and grade of VUR. This finding is important, since it suggests that VUR in itself reinforces the inflammatory response, and thereby has the potential to increase the risk of permanent kidney damage following UTI. Antibiotic prophylaxis has traditionally been used to prevent new infections in patients at risk. Despite our prescription of prophylactic drugs to children with dilating VUR (grades III to V), this group had significantly more febrile recurrences than those with low grade or no VUR. There is also a lack of controlled studies showing that prophylaxis is superior to prompt treatment of recurrent UTIs as far as renal outcome is concerned.15 Finally, there is a need for more information about which patients really benefit from long-term prophylaxis as well as the optimal duration of treatment. Whether active treatment of VUR hinders or inhibits development or progress of scarring has been debated for several years.16 In a meta-analysis of children with UTI VUR was a weak predictor of renal damage, although it was not possible to extract sufficient data to evaluate whether more severe grades of VUR were better predictors of renal damage compared to minor grades.17 In a 10-year followup study of patients with grades III and IV VUR randomized to
650
VESICOURETERAL REFLUX, INFECTION AND RENAL DAMAGE
either surgical or medical treatment there was no difference with regard to identification of new renal scars.18 New scars developed mostly in children younger than 5 years, and more frequently in those with grade IV than with grade III reflux. In a meta-analysis of randomized controlled trials comparing antibiotics and surgery for VUR it was not clear whether any intervention was beneficial.19 During recent years an endoscopic technique to treat VUR has been introduced.20 This procedure is less invasive and has replaced reimplantation at many centers. However, a major problem is that no known controlled study has been published evaluating the effect of this procedure on the frequency of febrile recurrences, or on the long-term outcome of kidney development or function. Thus, the indications for active treatment vs antibiotic prophylaxis need to be addressed in controlled randomized trials, which also should include a group managed by observation and prompt treatment of each UTI as it occurs. The Swedish Reflux Study is an ongoing project comparing these 3 alternatives, with febrile reinfections, renal damage on DMSA scan and VUR outcome as end points. CONCLUSIONS This study reveals an association between VUR grade and permanent renal damage in boys and girls. In girls the grade of inflammation at first UTI was also a significant factor, and one that might be even more important than reflux grade. The study lends support to the concept that renal damage is often congenital and is associated with VUR in boys, whereas in girls it is more often acquired and related to the severity of inflammatory reaction.
6. 7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
ACKNOWLEDGMENTS 17.
Gunnar Ekeroth and Emma Jaensson assisted with the statistical analysis.
Abbreviations and Acronyms CRP DMSA UTI VCU VUR
⫽ ⫽ ⫽ ⫽ ⫽
C-reactive protein dimercapto-succinic acid urinary tract infection voiding cystourethrography vesicoureteral reflux
REFERENCES 1.
Jakobsson B, Esbjorner E and Hansson S: Minimum incidence and diagnostic rate of first urinary tract infection. Pediatrics 1999; 104: 222. 2. Glauser MP, Meylan P and Bille J: The inflammatory response and tissue damage. The example of renal scars following acute renal infection. Pediatr Nephrol 1987; 1: 615. 3. Svanborg C, Bergsten G, Fischer H, Frendeus B, Godaly G, Gustafsson E et al: The ‘innate’ host response protects and damages the infected urinary tract. Ann Med 2001; 33: 563. 4. Stokland E, Hellstrom M, Jacobsson B, Jodal U and Sixt R: Renal damage one year after first urinary tract infection: role of dimercaptosuccinic acid scintigraphy. J Pediatr 1996; 129: 815. 5. Risdon RA: The small scarred kidney of childhood. A congenital or an acquired lesion? Pediatr Nephrol 1987; 1: 632.
18.
19.
20.
Risdon RA: The small scarred kidney in childhood. Pediatr Nephrol 1993; 7: 361. Yeung CK, Godley ML, Dhillon HK, Gordon I, Duffy PG and Ransley PG: The characteristics of primary vesico-ureteric reflux in male and female infants with pre-natal hydronephrosis. Br J Urol 1997; 80: 319. Wennerstrom M, Hansson S, Jodal U and Stokland E: Primary and acquired renal scarring in boys and girls with urinary tract infection. J Pediatr 2000; 136: 30. Hansson S, Dhamey M, Sigstrom O, Sixt R, Stokland E, Wennerstrom M et al: Dimercapto-succinic acid scintigraphy instead of voiding cystourethrography for infants with urinary tract infection. J Urol 2004; 172: 1071. Lebowitz RL, Olbing H, Parkkulainen KV, Smellie JM and Tamminen-Mobius TE: International system of radiographic grading of vesicoureteric reflux. International Reflux Study in Children. Pediatr Radiol 1985; 15: 105. Hansson S, Bollgren I, Esbjorner E, Jakobsson B and Marild S: Urinary tract infections in children below two years of age: a quality assurance project in Sweden. The Swedish Pediatric Nephrology Association. Acta Paediatr 1999; 88: 270. Jacobson SH, Hansson S and Jakobsson B: Vesico-ureteric reflux: occurrence and long-term risks. Acta Paediatr, suppl., 1999; 88: 22. Rushton HG: The evaluation of acute pyelonephritis and renal scarring with technetium 99m-dimercaptosuccinic acid renal scintigraphy: evolving concepts and future directions. Pediatr Nephrol 1997; 11: 108. Martinell J, Lidin-Janson G, Jagenburg R, Sivertsson R, Claesson I and Jodal U: Girls prone to urinary infections followed into adulthood. Indices of renal disease. Pediatr Nephrol 1996; 10: 139. Williams G, Lee A and Craig J: Antibiotics for the prevention of urinary tract infection in children: a systematic review of randomized controlled trials. J Pediatr 2001; 138: 868. Jodal U, Hansson S and Hjalmas K: Medical or surgical management for children with vesico-ureteric reflux? Acta Paediatr, suppl., 1999; 88: 53. Gordon I, Barkovics M, Pindoria S, Cole T and Woolf AS: Primary vesicoureteric reflux as a predictor of renal damage in children hospitalized with urinary tract infection: a systematic review and meta-analysis. J Am Soc Nephrol 2003; 14: 739. Olbing H, Smellie JM, Jodal U and Lax H: New renal scars in children with severe VUR: a 10-year study of randomized treatment. Pediatr Nephrol 2003; 18: 1128. Wheeler D, Vimalachandra D, Hodson EM, Roy LP, Smith G and Craig JC: Antibiotics and surgery for vesicoureteric reflux: a meta-analysis of randomised controlled trials. Arch Dis Child 2003; 88: 688. Lackgren G, Wahlin N and Stenberg A: Endoscopic treatment of children with vesico-ureteric reflux. Acta Paediatr, suppl., 1999; 88: 62.
EDITORIAL COMMENTS The specific aim of the authors was to study the relationship between urinary tract infection, vesicoureteral reflux and acquired renal cortical scarring in a retrospective analysis of 303 children younger than 2 years with a first time, culture documented, symptomatic urinary tract infection. Each patient underwent DMSA renal scintigraphy within 3 months of presenting for UTI, as well as after 1 to 2 years of rigorous followup while on antibiotic prophylaxis. The patients studied represent the same cohort that formed the basis of an earlier series supporting the use of DMSA renal scanning as the first imaging tool in the evaluation of UTI because of its ability to indicate the presence of vesicoureteral reflux based
VESICOURETERAL REFLUX, INFECTION AND RENAL DAMAGE on abnormalities seen in the renal parenchyma consistent with acute pyelonephritis. In that study 156 of the 303 children (51%) presenting with a culture proved UTI and fever had an abnormal DMSA scan initially, thus, excluding the remaining children from being considered as having radiologically proved acute pyelonephritis. The authors found a significant relationship between the presence of acute lesions and the presence and severity of VUR (p ⬍0.001), and in particular there was an increased risk of DMSA scan abnormalities in boys and girls with grade III and higher VUR (reference 9 in article). In the current study 26% of children presenting with symptomatic UTI had DMSA renal scan abnormalities on followup 2 years later. It is noteworthy that 49% of the children were unlikely to form scars simply on the basis of negative DMSA renal scan abnormalities on initial presentation. Taking this finding into account, it can be estimated that the incidence of renal scarring in children with initially positive renal scans is 80 of 156, or 51%, which is much higher than what has been reported in other studies. Thus, this patient population is heterogeneous (includes patients without radiographically proved acute pyelonephritis) and may include a cohort of children with a greater predisposition to renal scarring following pyelonephritis than is typically seen. The authors found that DMSA renal scan abnormalities were correlated with the presence of VUR grade II and higher in children of both genders, with maximum recorded body temperature and maximum C-reactive protein level at first UTI, and with the number of recurrent UTIs while on prophylaxis. However, in a stepwise logistic regression analysis searching for independent variables only VUR was strongly correlated with renal scarring, with the strongest correlation occurring in boys. The relative risk that VUR was associated with renal scan abnormalities increased significantly by grade, from 1.20 for grade I to 4.61 for grades IV and V. It is noteworthy that boys were more likely to have higher grades of reflux, a finding that raises suspicion that some of the abnormalities noted at 2 years might in fact have been present at birth. An additional interesting observation can be made in the cohort of children (15 of 80) who had dilating vesicoureteral reflux (grades III to IV) and initially positive DMSA scans
651
(including 2 children who also had recurrent UTI), who were obviously at low risk for renal scarring on the basis of normal followup DMSA renal scans. Clearly, some children are inherently protected from the ill effects of acute pyelonephritis by mechanisms that are as yet unclear to us. Hans G. Pohl Department of Pediatric Urology George Washington University Children’s National Medical Center Washington, D. C.
The authors seem to have shown that there is an association between reflux and recurrent urinary tract infection, and reflux and abnormal nuclear renal scans. Of the children 12% had a febrile UTI recurrence, and this outcome correlated with the presence and severity of reflux. Abnormalities in the renal scan occurred more often in patients who had reflux. However, the question remains as to whether these findings are due to cause and effect, or are merely associations. Reading this article, one is interested in the CRP data. In the population studied CRP levels correlated with DMSA abnormalities. Analyzing this data, it would seem that girls with a CRP of less than 50 and boys with a CRP of less than 20 should not receive a DMSA scan, since the chance of an abnormality on this study at those CRP levels is quite low. The data presented in figure 1 led to another question. If these data were presented in a scatter plot, would there be a CRP level low enough that could confidently correlate with absolutely no cases of reflux? If so, many children might be spared the unpopular voiding cystourethrogram. All in all, I believe that the authors do a service to our specialty and to pediatricians at large because their findings countermand some of the growing voices that suggest urinary reflux is not a significant condition. Richard N. Schlussel Department of Urology Columbia University Medical Center New York, New York