- Email: [email protected]
Renal Scars and Parenchymal Thinning in Children with Vesicoureteral Reflux: A 5-Year Report of the International Reflux Study in Children (European Branch)
0022-5347 /92/14.85--1653$03,00/0 'fol. 148, 1653-1656,
THE JOURNAL OF UROLOGY
November 1992
Printed in U.S.A.
Copyright© 1992 by AMERICAN UROLOGICAL ASSOCIATION, INC,
RENAL SCARS AND PARENCHYMAL THINNING IN CHILDREN ,vITH VESICOURETERAL REFLUX: A 5-YEAR REPORT OF THE INTERNATIONAL REFLUX STUDY IN CHILDREN (EUROPEAN BRANCH) H. OLBING, I. CLAESSON, K.-D. EBEL, U. SEPPANEN, J. M. SMELLIE, T. TAMMINEN-MOBIUS AND I. WIKST AD ON BEHALF OF THE INTERNATIONAL REFLUX STUDY IN CHILDREN
ABSTRACT
A total of 321 children less than 11 years old with nonobstructive grade III or IV vesicoureteral reflux and with previous urinary tract infection was randomly allocated to medical or surgical treatment in the European branch of the International Reflux Study in Children. (Randomization was stratified for age, sex, grade of reflux, presence of renal scarring, interval since last urinary tract infection and treating hospital.) The results of excretory urography are reported for 233 girls and 73 boys treated according to the random allocation, 89% of whom were followed for 5 years. After 5 years in the medical group (155 children) new renal scars were seen in 19 and new renal parenchymal thinning in 11. The proportions were almost identical among 151 children allocated to surgical treatment with 20 new scars and 15 new thinnings. Progression of established scars was also similar in both groups. However, the new scars developed sooner after surgery than during medical treatment. In 6 surgically treated children postoperative obstruction was followed by the development of new scars. In addition, 12 patients showed new scars approximately 6 months after successful surgery, while in only 2 children scars developed more than 6 months after surgery. In 11 children of the medical group new scars were seen more than 6 months after allocation. More new scars developed in the children with parenchymal thinning at entry (23 %) than in those with scarred or normal kidneys at entry (10% each) (p <0.05). The younger the patients at entry, the higher the frequency of new scars (less than 2 years 19.8%) 2 to 4 years 9.8% and 5 years or more 4.6%, p <0.05). Renal scars are the main renal lesions associated with vesicoureteral reflux, and their radiological appearance on excretory urography (IVP) has been described by Hodson. 1 In the absence of outflow obstruction, the presence of vesicoureteral reflux, patent papillary duct collecting orifices and urinary tract infection is necessary for the development of renal scars. 2 • 3 In view of these facts, surgical correction of vesicoureteral reflux or prevention of urinary tract infection recurrences mainly by low dose antibacterial prophylaxis has been used worldwide in children with nonobstructive reflux. Thinning of renal parenchyma in children with vesicoureteral reflux has occasionally been described, 4- 11 The pathogenesis and the future course of acquired parenchymal thinning have not yet been studied. We designed a prospective multicenter study comparing medical and surgical management of children with nonobstructive Participating hospital§ and principal investigators: University Children's Hospital, Bonn, Germany: Rudolf Mallmann. University Children's Hospital, Universite Libre de Bruxelles, Belgium: Michelle Hall. University Children's Hospital, Essen, Germany (Coordinating Center): Hermann Olbing (Chairman), Tytti Tamminen-Miibius (Coordinator) and Wolfgang Rascher. University Children's Hospital, Gothenburg, Sweden: Kelm Hjiilmas and Ulf JodaL Department of Urology, University of Hamburg, Germany: Rainer Busch. University Children's Hospital, Helsinki, Finland: Olli Koskimies. University Children's Hospital, Oulu, Finland: Juhani Seppanen. Karolinska Institutet, St. Goran's and Sachska Children's Hospitals, Stockholm, Sweden: Anita Aperia. Data processing and statistical analysis: Institute for Medical Data Processing and Biomathemathics, University Essen, Germany: Hildegard Lax-Gro/3 and Herbert Hirche. Scientific advisors: Klaus-Dieter Ebel (pediatric radiology), Jan van Goo! (pediatric urodynamics), Kalle V. Parkkulainen (pediatric surgery), Jean M. Smellie (pediatrics) and Jan Winberg (pediatrics) Supported by the VW-Foundation (Grants AZ 35 807, AZ I/37 504) until 1985 and thereafter by the Bundesministerium fur Forschung und Technologie (Grant 07068343).
vesicoureteral reflux with dilatation. 12 We report the results of a 5-year followup of 233 girls and 73 boys from 8 European participating hospitals. At the same time 132 comparable children from 16 American centers were studied with similar methods and the results will be published separately. METHODS
IVP was scheduled for each patient at entry into the study and 6 months, 18 months and 54 months after entry or surgery. No IVP was performed at entry in those children for whom an adequate IVP performed not more than 6 months previously was available. Four children had no IVP followup, and 6 had the last IVP performed after 6 months, 15 after 18 months and 9 after 30 or 42 months. Thinning of renal parenchyma was defined as localized or generalized reduction of parenchymal thickness (at !east -2.5 standard deviations) with normal underlying calices (fig. 1, A). Hodson's definition of renal scars was used (fig. 1, B) 1 and scars were typed according to Smellie et al. B IVP method, definition of scar and scar types are described elsewhere. 14 All IVPs were evaluated by at least 3 experts, and those considered abnormal were reviewed by a panel of 5. Two renal tracings were made independently and compared; if they agreed parenchymal thickness of the upper pole, lower pole, lateral zone and projected planimetric area was measured. The results were related to the height of 11-13 and compared with a nomogram. 15• 16 Progression of established renal scars could be either local or generalized. Local progression was defined as the combination of an increase in caliceal deformity (widening and rounding) and further reduction of the overlying parenchyma without change of projected planimetric area. If the increase of
1653
1654
OLBING AND ASSOCIATES
FIG. 1. Consecutive IVPs of 3-year-old girl. A, thin renal parenchyma overlying normal calix in right upper pole (-3 standard deviations) at entry. B, renal scar in upper pole (thin parenchyma overlying clubbed calix) 12 months later.
caliceal deformity involved several scarred segments and projected planimetric area decreased by more than 1 standard deviation the progression was classified as general. A life table analysis according to Kaplan-Meier with generalized rank test of Gehan-Wilcoxon was performed to compare both groups in the primary outcome regarding the development or progression of renal scars. Because stratified randomization created 2 groups with similar distribution of patient entry characteristics, it was possible to test the frequency tables comparing the groups with univariate chi-square tests according to Brand/Snedecor or the chi-square trend test of Cochran.
1. Age and reflux grades at entry, and sex of 306 children of
TABLE
different treatment groups No. Children Medical Sex: M
F Age (yrs.): Less than 2 2-4 5 or more Reflux grade*:
III IV Total No. pts.
PATIENTS
A total of 401 children less than 11 years old with nonobstructive grade III or IV vesicoureteral reflux (international classification), a history of urinary tract infection and a glomerular filtration rate of at least 70 ml. per minute per 1. 73 m. 2 was recruited from January 1980 to March 1985 and managed on a medical regimen (continuous low dose antibiotic prophylaxis and attention to bowel function and regular micturition). Infants were accepted only with grade IV vesicoureteral reflux and a glomerular filtration rate of at least 60% of the normal value for age. The 321 patients with persisting grade III or IV vesicoureteral reflux after a run-in observation period of 2 to 7 months were randomly allocated to surgical or to continued medical management, stratified for sex, age, reflux grade, renal scars at entry, interval since last urinary tract infection and the different participating hospitals. Of these children 306 were treated according to the random allocation of whom 89% were followed for up to 5 years. Table 1 shows the sex, age and grade of reflux at entry to the study of the 306 European children treated medically or surgically according to random allocation. Of the children 49% already had renal scars, 17% showed areas of renal parenchymal thinning without scars and 34% had normal IVPs. The distribution of different types of scars and thin areas was not significantly different between the 2 treatment groups (table 2). Further details of patient characteristics at entry are published in another article. 14 The history of urinary tract infection was a stratification criterion (interval since latest urinary tract infection at entry). There was no difference in the frequency
Surgical
Totals
37 118
36 115
73 233
50 56 49
47 66 38
97 122 87
31 (23) 124 (96) 155 (119)
25 (19) 126 (99) 151 (118)
56 (42) 250 (195) 306 (237)
* The number of children with bilateral reflux is given in parentheses based on the higher grade. TABLE
2. Kidney state on IVP at entry in children in medical and
surgical treatment groups No. Children Kidney State Medical Normal Scars Scar types: A B C
D Thin areas Totals
Surgical 51 70
54 79 30 (5) 32 (13) 10 (7) 7 (2) 22 (2) 155
25 32 8 5 30 151
Totals 105 149
(5) (12) (2) (0) (3)
55 (10) 64 (25) 18 (9) 12 (2) 52 (5) 306
Numbers in parentheses indicate 46 had bilateral scars, recorded according to the more severely scarred kidney. A further 9 medical and 8 surgical cases with scars had parenchymal thinning in the contralateral kidney. Thin parenchyma was found bilaterally in 2 medical and 3 surgical cases.
and type of urinary tract infections, and the interval between latest infection and entry was similar between the 2 groups. 17 RESULTS
New renal scars and areas of parenchymal thinning. New renal scars developed in 39 of the 306 children during the 5year observation, and new areas of renal parenchymal thinning developed in another 26. The proportion of patients with new
1655
RENAL SCARS AND PARENCHYMAL THINNING IN CHILDREN WITH VESICOURETERAL REFLUX
scars and with new thin areas did not differ between medically and surgically treated children (table 3). A life table analysis starting with the date of entry IVP showed no significant difference between the treatment groups (fig. 2). However, the interval from randomization to first documentation of the scar differed slightly in the 2 treatment groups. New scars were observed at the first followup IVP (which was planned 6 months after randomization or surgery) of 11 of the 19 medically treated and 18 of the 20 surgically treated children with new scars. Further scarring or new scars later than 6 months after entry were observed in 11 medically treated children, of whom 9 had persistent vesicoureteral reflux. New scars developed later than 6 months postoperatively in 2 children with persistent vesicoureteral reflux. Regardless of sex, there was a reciprocal correlation between TABLE
3. Changes of kidney state on IVP within 5 years in children
of different treatment groups No. Children Changes of Kidney State Medical New scars at entry: Normal Thin areas Scars New thin areas at entry: Normal Thin areas Scars Progression of established scars: Local General Thin areas no longer seen No change at entry: Normal Thinnings Scars Missing information Totals
Surgical
Totals
5 7 7
7 (2) 5 (1) 8 (3)
12 12 15
3 3 5
9 (2) 4 2
12 7 7
16 17 4
13 17 5
29 34 9
47
37 16 26 2 151
84 26 55 4 306
10
29 2 155
Numbers in parentheses indicate association with postoperative obstruction.
30
Different types of new renal scars on IVP observed during 5 years in children of different treatment groups No. Children
Types of New Scars A
B C D Totals
Medical
Surgical
8 (6) 8 (4) 2 (1) 1 19 (11)
8 5 7* 0 20
Totals
(7) (2) (3)
16 13 9 1
(12)
39
Numbers in parentheses indicate children with normal kidneys or kidneys with thin parenchyma at entry. * An additional patient (counted as B) had transient type C kidney with change to type B 4 years after release of obstruction.
age at entry and the frequency of new scars. New scars developed in 23. 7% among those who were less than 2 years old at study entry, 9.8% of those who were 2 to 4 years old and 4.6% of those who were 5 years or older (p <0.05). New scars were more common in boys (18%) than girls (11%). However, this difference was not statistically significant (p <0.1) and may be related to the higher proportion of boys than girls entering the study less than age 2 years. Among the children in whom new scars developed the numbers with types A, B and D did not differ between the medical and surgical groups. Of the 8 surgical patients in whom type C scars developed 6 had been exposed to postoperative obstruction (table 4). 18 Changes in children with thin renal parenchymal areas at entry and progression of renal scars. Among the 52 children who had thin areas without scars at entry new scars developed in the area that was previously thin in 23%, additional thin areas developed in 12% and the thin areas were no longer apparent after 5 years in 10%. By comparison, new scars developed in only 10% of the children who entered the study with normal or scarred kidneys (p <0.01). Among the 149 children with renal scars at entry into the study, the scars progressed in 63 (29 locally and 34 generalized). Again, no difference was demonstrable between the medical and the surgical groups (table 3). DISCUSSION
149
126
116 N Surg.
153
141
121 N Med.
10
5 0
TABLE 4.
~-----..---------......--.0 10 20 30 40 50 60 Months
-surgical •••Med I ca I
{n=151) (n=155)
FIG. 2. Acquisition of new renal scars during 5 years in surgical and medical treatment groups. Life table analysis showed no difference between groups after 5-year followup. 0, date of entry IVP. Followup IVPs were planned for 6, 18 and 54 months after randomization or surgery.
We report the findings on IVP at entry and within 5 years of the European branch of an international randomized treatment (medical or surgical) trial of children with nonobstructive grade III or IV vesicoureteral reflux and a history of urinary tract infection. The 2 treatment groups were similar in terms of age, sex, reflux grade, presence of renal scars at entry, interval since last urinary tract infection, frequency of previous urinary tract infection (including pyelonephritic episodes) and patient numbers for each of the different treatment centers. After 5 years there was no difference in the frequency of new renal scars, new areas of renal parenchymal thinning or progression of established scars between the children treated medically or surgically. This finding confirms with a greater number of patients the results of the Birmingham Reflux Study Group that surgical and medical treatment can be equally effective in the nrevention of new renal scar areas in a 5-year period. 19• 20 Although similar numbers of new scars were observed in the 2 treatment groups at the end of a 5-year observation period, the new scars were recognized somewhat earlier in the surgical than in the medical group. Among the possible causes, postoperative obstruction, which was documented in 10 of the 237 ureters (10 of 151 surgical patients), 19 must be considered. Surgically and medically treated children had been exposed to similar numbers of pyelonephritic episodes and they had similar urinary tract infection-free intervals before entry. 17 The higher frequency of pyelonephritic attacks during the study in the medical com pared to the surgical group 17 might explain the later development of some new scars in the medical group. Of the 52 children with areas of thin renal parenchyma at
1656
OLBING AND ASSOCIATES
entry into the study new scars developed in the same area in 12 (23%) compared to 27 children with new scars among the other 254 patients in the study (10.6%). It may be concluded that thinning of renal parenchyma can be a potential precursor of a renal scar. In 9 of the 52 children the thinning of renal parenchyma seen at entry was no longer visible after 5 years. Thus, renal thinning may be potentially reversible. Renal parenchymal thinning observed at entry into the study and not changing during 5 years may constitute normal morphological variation or arrested renal damage. REFERENCES 1. Hodson, C. J.: The role of radiological diagnosis. Proc. Roy. Soc. Med., 52: 669, 1959. ~- Hodson, C. J., Maling, T. M.-J:, McManamon;P.J. and Lewis, M. G.: The pathogenesis of reflux nephropathy (chronic atrophic pyelonephritis). Brit. J. Rad., suppl., 13: 1, 1975. 3. Ransley, P. G. and Risdon, R. A.: Reflux and renal scarring. Brit. J. Rad., suppl., 14: 1, 1978. 4. Hodson, C. J. and Edwards, D.: Chronic pyelonephritis and vesicoureteric reflux. Clin. Rad., 11: 219, 1960. 5. Babcock, J. R., Keats, G. K. and King, L. R.: Renal changes after an uncomplicated antireflux operation. J. Urol., 115: 720, 1976. 6. Winberg, J., Claesson, I., Jacobson, B., Jodal, U. and Peterson, H.: Renal growth after acute pyelonephritis in childhood: an epidemiological approach. In: Reflux Nephropathy. Edited by J. Hodson and P. Kincaid-Smith. New York: Masson Publishers, pp. 309-322, 1979. 7. Claesson, I., Jacobsson, B., Jodal, U. and Winberg, J.: Early detection of nephropathy in childhood urinary tract infection. Acta Rad. (Diagn.), 22: 315, 1981. 8. Tamminen, T., Olbing, H. and Bachmann, H. J.: The development of segmental scarring in previously normal kidneys seen in three children with vesicoureteric reflux. Klin. Piidiat., 194: 137, 1982. 9. Hellstrom, M., Jacobsson, B., Jodal, U., Winberg, J. and Oden, A.:
10. 11. 12. 13. 14.
15. 16. 17.
18.
19. 20.
Renal growth after neonatal urinary tract infection. Ped. Nephrol., 1: 269, 1987. Harrison, B. R., Howards, S. S. and Thomas, R.: Medical deviation of the upper pole calix on the intravenous urogram as an indication of vesico-ureteric reflux. AJR, 126: 1189, 1976. Gedroyc, W. M. W., Chaudhuri, R. and Saxton, H. M.: Normal and near normal caliceal patterns in reflux nephropathy. Clin. Rad., 39: 615, 1988. International Reflux Study Committee: Medical versus surgical treatment of primary vesico-ureteral reflux: a prospective international study in children. J. Urol., 125: 244, 1981. Smellie, J., Edwards, D., Hunter, N., Normand, I. C. S. and Prescod, N.: Vesico-ureteric reflux and renal scarring. Kidney Int., 8: 65, 1975. Weiss, R., Tamminen-Mobius, T., Koskimies, 0., Olbing, H., Smellie, J. M., Hirche, H. and Lax-Gross, H.: Characteristics at entry of children with severe primary vesicoureteral reflux recruited for a multicente-r, international therapeutic trial eomparing medical and surgical management. J. Urol., part 2, 148: 1644, 1992. Jorulf, H., Nordmark, J. and Jonsson, A.: Kidney size in infants and children assessed by area measurement. Acta Rad. (Diagn.), 19: 154, 1978. Claesson, I., Jacobsson, B., Olsson, T. and Ringertz, H.: Assessment of renal parenchymal thickness in normal children. Acta Rad. (Diagn.), 22: 305, 1981. Jodal, U., Koskimies, 0., Hanson, E., Lohr, G., Olbing, H., Smellie, J. and Tamminen-Mobius, T.: Infection pattern in children with vesicoureteral reflux randomly allocated to operation or longterm antibacterial prophylaxis. J. Urol., part 2, 148: 1650, 1992. Hjiilmas, K., Lohr, G., Tamminen-Mobius, T., Seppanen, J., Olbing, H. and Wikstrom, S.: Surgical results in the International Reflux Study in Children (Europe). J. Urol., part 2, 148: 1657, 1992. Birmingham Reflux Study Group: Prospective trial of operative versus non-operative treatment of severe vesico-ureteric reflux: 2 years' observation in 96 children. Brit. Med. J., 287: 171, 1983. Birmingham Reflux Study Group: Prospective trial of operative versus non-operative treatment of severe vesicoureteric reflux in children: five years' observation. Brit. Med. J., 295: 237, 1987.
THE JOURNAL OF UROLOGY
November 1992
Printed in U.S.A.
Copyright© 1992 by AMERICAN UROLOGICAL ASSOCIATION, INC,
RENAL SCARS AND PARENCHYMAL THINNING IN CHILDREN ,vITH VESICOURETERAL REFLUX: A 5-YEAR REPORT OF THE INTERNATIONAL REFLUX STUDY IN CHILDREN (EUROPEAN BRANCH) H. OLBING, I. CLAESSON, K.-D. EBEL, U. SEPPANEN, J. M. SMELLIE, T. TAMMINEN-MOBIUS AND I. WIKST AD ON BEHALF OF THE INTERNATIONAL REFLUX STUDY IN CHILDREN
ABSTRACT
A total of 321 children less than 11 years old with nonobstructive grade III or IV vesicoureteral reflux and with previous urinary tract infection was randomly allocated to medical or surgical treatment in the European branch of the International Reflux Study in Children. (Randomization was stratified for age, sex, grade of reflux, presence of renal scarring, interval since last urinary tract infection and treating hospital.) The results of excretory urography are reported for 233 girls and 73 boys treated according to the random allocation, 89% of whom were followed for 5 years. After 5 years in the medical group (155 children) new renal scars were seen in 19 and new renal parenchymal thinning in 11. The proportions were almost identical among 151 children allocated to surgical treatment with 20 new scars and 15 new thinnings. Progression of established scars was also similar in both groups. However, the new scars developed sooner after surgery than during medical treatment. In 6 surgically treated children postoperative obstruction was followed by the development of new scars. In addition, 12 patients showed new scars approximately 6 months after successful surgery, while in only 2 children scars developed more than 6 months after surgery. In 11 children of the medical group new scars were seen more than 6 months after allocation. More new scars developed in the children with parenchymal thinning at entry (23 %) than in those with scarred or normal kidneys at entry (10% each) (p <0.05). The younger the patients at entry, the higher the frequency of new scars (less than 2 years 19.8%) 2 to 4 years 9.8% and 5 years or more 4.6%, p <0.05). Renal scars are the main renal lesions associated with vesicoureteral reflux, and their radiological appearance on excretory urography (IVP) has been described by Hodson. 1 In the absence of outflow obstruction, the presence of vesicoureteral reflux, patent papillary duct collecting orifices and urinary tract infection is necessary for the development of renal scars. 2 • 3 In view of these facts, surgical correction of vesicoureteral reflux or prevention of urinary tract infection recurrences mainly by low dose antibacterial prophylaxis has been used worldwide in children with nonobstructive reflux. Thinning of renal parenchyma in children with vesicoureteral reflux has occasionally been described, 4- 11 The pathogenesis and the future course of acquired parenchymal thinning have not yet been studied. We designed a prospective multicenter study comparing medical and surgical management of children with nonobstructive Participating hospital§ and principal investigators: University Children's Hospital, Bonn, Germany: Rudolf Mallmann. University Children's Hospital, Universite Libre de Bruxelles, Belgium: Michelle Hall. University Children's Hospital, Essen, Germany (Coordinating Center): Hermann Olbing (Chairman), Tytti Tamminen-Miibius (Coordinator) and Wolfgang Rascher. University Children's Hospital, Gothenburg, Sweden: Kelm Hjiilmas and Ulf JodaL Department of Urology, University of Hamburg, Germany: Rainer Busch. University Children's Hospital, Helsinki, Finland: Olli Koskimies. University Children's Hospital, Oulu, Finland: Juhani Seppanen. Karolinska Institutet, St. Goran's and Sachska Children's Hospitals, Stockholm, Sweden: Anita Aperia. Data processing and statistical analysis: Institute for Medical Data Processing and Biomathemathics, University Essen, Germany: Hildegard Lax-Gro/3 and Herbert Hirche. Scientific advisors: Klaus-Dieter Ebel (pediatric radiology), Jan van Goo! (pediatric urodynamics), Kalle V. Parkkulainen (pediatric surgery), Jean M. Smellie (pediatrics) and Jan Winberg (pediatrics) Supported by the VW-Foundation (Grants AZ 35 807, AZ I/37 504) until 1985 and thereafter by the Bundesministerium fur Forschung und Technologie (Grant 07068343).
vesicoureteral reflux with dilatation. 12 We report the results of a 5-year followup of 233 girls and 73 boys from 8 European participating hospitals. At the same time 132 comparable children from 16 American centers were studied with similar methods and the results will be published separately. METHODS
IVP was scheduled for each patient at entry into the study and 6 months, 18 months and 54 months after entry or surgery. No IVP was performed at entry in those children for whom an adequate IVP performed not more than 6 months previously was available. Four children had no IVP followup, and 6 had the last IVP performed after 6 months, 15 after 18 months and 9 after 30 or 42 months. Thinning of renal parenchyma was defined as localized or generalized reduction of parenchymal thickness (at !east -2.5 standard deviations) with normal underlying calices (fig. 1, A). Hodson's definition of renal scars was used (fig. 1, B) 1 and scars were typed according to Smellie et al. B IVP method, definition of scar and scar types are described elsewhere. 14 All IVPs were evaluated by at least 3 experts, and those considered abnormal were reviewed by a panel of 5. Two renal tracings were made independently and compared; if they agreed parenchymal thickness of the upper pole, lower pole, lateral zone and projected planimetric area was measured. The results were related to the height of 11-13 and compared with a nomogram. 15• 16 Progression of established renal scars could be either local or generalized. Local progression was defined as the combination of an increase in caliceal deformity (widening and rounding) and further reduction of the overlying parenchyma without change of projected planimetric area. If the increase of
1653
1654
OLBING AND ASSOCIATES
FIG. 1. Consecutive IVPs of 3-year-old girl. A, thin renal parenchyma overlying normal calix in right upper pole (-3 standard deviations) at entry. B, renal scar in upper pole (thin parenchyma overlying clubbed calix) 12 months later.
caliceal deformity involved several scarred segments and projected planimetric area decreased by more than 1 standard deviation the progression was classified as general. A life table analysis according to Kaplan-Meier with generalized rank test of Gehan-Wilcoxon was performed to compare both groups in the primary outcome regarding the development or progression of renal scars. Because stratified randomization created 2 groups with similar distribution of patient entry characteristics, it was possible to test the frequency tables comparing the groups with univariate chi-square tests according to Brand/Snedecor or the chi-square trend test of Cochran.
1. Age and reflux grades at entry, and sex of 306 children of
TABLE
different treatment groups No. Children Medical Sex: M
F Age (yrs.): Less than 2 2-4 5 or more Reflux grade*:
III IV Total No. pts.
PATIENTS
A total of 401 children less than 11 years old with nonobstructive grade III or IV vesicoureteral reflux (international classification), a history of urinary tract infection and a glomerular filtration rate of at least 70 ml. per minute per 1. 73 m. 2 was recruited from January 1980 to March 1985 and managed on a medical regimen (continuous low dose antibiotic prophylaxis and attention to bowel function and regular micturition). Infants were accepted only with grade IV vesicoureteral reflux and a glomerular filtration rate of at least 60% of the normal value for age. The 321 patients with persisting grade III or IV vesicoureteral reflux after a run-in observation period of 2 to 7 months were randomly allocated to surgical or to continued medical management, stratified for sex, age, reflux grade, renal scars at entry, interval since last urinary tract infection and the different participating hospitals. Of these children 306 were treated according to the random allocation of whom 89% were followed for up to 5 years. Table 1 shows the sex, age and grade of reflux at entry to the study of the 306 European children treated medically or surgically according to random allocation. Of the children 49% already had renal scars, 17% showed areas of renal parenchymal thinning without scars and 34% had normal IVPs. The distribution of different types of scars and thin areas was not significantly different between the 2 treatment groups (table 2). Further details of patient characteristics at entry are published in another article. 14 The history of urinary tract infection was a stratification criterion (interval since latest urinary tract infection at entry). There was no difference in the frequency
Surgical
Totals
37 118
36 115
73 233
50 56 49
47 66 38
97 122 87
31 (23) 124 (96) 155 (119)
25 (19) 126 (99) 151 (118)
56 (42) 250 (195) 306 (237)
* The number of children with bilateral reflux is given in parentheses based on the higher grade. TABLE
2. Kidney state on IVP at entry in children in medical and
surgical treatment groups No. Children Kidney State Medical Normal Scars Scar types: A B C
D Thin areas Totals
Surgical 51 70
54 79 30 (5) 32 (13) 10 (7) 7 (2) 22 (2) 155
25 32 8 5 30 151
Totals 105 149
(5) (12) (2) (0) (3)
55 (10) 64 (25) 18 (9) 12 (2) 52 (5) 306
Numbers in parentheses indicate 46 had bilateral scars, recorded according to the more severely scarred kidney. A further 9 medical and 8 surgical cases with scars had parenchymal thinning in the contralateral kidney. Thin parenchyma was found bilaterally in 2 medical and 3 surgical cases.
and type of urinary tract infections, and the interval between latest infection and entry was similar between the 2 groups. 17 RESULTS
New renal scars and areas of parenchymal thinning. New renal scars developed in 39 of the 306 children during the 5year observation, and new areas of renal parenchymal thinning developed in another 26. The proportion of patients with new
1655
RENAL SCARS AND PARENCHYMAL THINNING IN CHILDREN WITH VESICOURETERAL REFLUX
scars and with new thin areas did not differ between medically and surgically treated children (table 3). A life table analysis starting with the date of entry IVP showed no significant difference between the treatment groups (fig. 2). However, the interval from randomization to first documentation of the scar differed slightly in the 2 treatment groups. New scars were observed at the first followup IVP (which was planned 6 months after randomization or surgery) of 11 of the 19 medically treated and 18 of the 20 surgically treated children with new scars. Further scarring or new scars later than 6 months after entry were observed in 11 medically treated children, of whom 9 had persistent vesicoureteral reflux. New scars developed later than 6 months postoperatively in 2 children with persistent vesicoureteral reflux. Regardless of sex, there was a reciprocal correlation between TABLE
3. Changes of kidney state on IVP within 5 years in children
of different treatment groups No. Children Changes of Kidney State Medical New scars at entry: Normal Thin areas Scars New thin areas at entry: Normal Thin areas Scars Progression of established scars: Local General Thin areas no longer seen No change at entry: Normal Thinnings Scars Missing information Totals
Surgical
Totals
5 7 7
7 (2) 5 (1) 8 (3)
12 12 15
3 3 5
9 (2) 4 2
12 7 7
16 17 4
13 17 5
29 34 9
47
37 16 26 2 151
84 26 55 4 306
10
29 2 155
Numbers in parentheses indicate association with postoperative obstruction.
30
Different types of new renal scars on IVP observed during 5 years in children of different treatment groups No. Children
Types of New Scars A
B C D Totals
Medical
Surgical
8 (6) 8 (4) 2 (1) 1 19 (11)
8 5 7* 0 20
Totals
(7) (2) (3)
16 13 9 1
(12)
39
Numbers in parentheses indicate children with normal kidneys or kidneys with thin parenchyma at entry. * An additional patient (counted as B) had transient type C kidney with change to type B 4 years after release of obstruction.
age at entry and the frequency of new scars. New scars developed in 23. 7% among those who were less than 2 years old at study entry, 9.8% of those who were 2 to 4 years old and 4.6% of those who were 5 years or older (p <0.05). New scars were more common in boys (18%) than girls (11%). However, this difference was not statistically significant (p <0.1) and may be related to the higher proportion of boys than girls entering the study less than age 2 years. Among the children in whom new scars developed the numbers with types A, B and D did not differ between the medical and surgical groups. Of the 8 surgical patients in whom type C scars developed 6 had been exposed to postoperative obstruction (table 4). 18 Changes in children with thin renal parenchymal areas at entry and progression of renal scars. Among the 52 children who had thin areas without scars at entry new scars developed in the area that was previously thin in 23%, additional thin areas developed in 12% and the thin areas were no longer apparent after 5 years in 10%. By comparison, new scars developed in only 10% of the children who entered the study with normal or scarred kidneys (p <0.01). Among the 149 children with renal scars at entry into the study, the scars progressed in 63 (29 locally and 34 generalized). Again, no difference was demonstrable between the medical and the surgical groups (table 3). DISCUSSION
149
126
116 N Surg.
153
141
121 N Med.
10
5 0
TABLE 4.
~-----..---------......--.0 10 20 30 40 50 60 Months
-surgical •••Med I ca I
{n=151) (n=155)
FIG. 2. Acquisition of new renal scars during 5 years in surgical and medical treatment groups. Life table analysis showed no difference between groups after 5-year followup. 0, date of entry IVP. Followup IVPs were planned for 6, 18 and 54 months after randomization or surgery.
We report the findings on IVP at entry and within 5 years of the European branch of an international randomized treatment (medical or surgical) trial of children with nonobstructive grade III or IV vesicoureteral reflux and a history of urinary tract infection. The 2 treatment groups were similar in terms of age, sex, reflux grade, presence of renal scars at entry, interval since last urinary tract infection, frequency of previous urinary tract infection (including pyelonephritic episodes) and patient numbers for each of the different treatment centers. After 5 years there was no difference in the frequency of new renal scars, new areas of renal parenchymal thinning or progression of established scars between the children treated medically or surgically. This finding confirms with a greater number of patients the results of the Birmingham Reflux Study Group that surgical and medical treatment can be equally effective in the nrevention of new renal scar areas in a 5-year period. 19• 20 Although similar numbers of new scars were observed in the 2 treatment groups at the end of a 5-year observation period, the new scars were recognized somewhat earlier in the surgical than in the medical group. Among the possible causes, postoperative obstruction, which was documented in 10 of the 237 ureters (10 of 151 surgical patients), 19 must be considered. Surgically and medically treated children had been exposed to similar numbers of pyelonephritic episodes and they had similar urinary tract infection-free intervals before entry. 17 The higher frequency of pyelonephritic attacks during the study in the medical com pared to the surgical group 17 might explain the later development of some new scars in the medical group. Of the 52 children with areas of thin renal parenchyma at
1656
OLBING AND ASSOCIATES
entry into the study new scars developed in the same area in 12 (23%) compared to 27 children with new scars among the other 254 patients in the study (10.6%). It may be concluded that thinning of renal parenchyma can be a potential precursor of a renal scar. In 9 of the 52 children the thinning of renal parenchyma seen at entry was no longer visible after 5 years. Thus, renal thinning may be potentially reversible. Renal parenchymal thinning observed at entry into the study and not changing during 5 years may constitute normal morphological variation or arrested renal damage. REFERENCES 1. Hodson, C. J.: The role of radiological diagnosis. Proc. Roy. Soc. Med., 52: 669, 1959. ~- Hodson, C. J., Maling, T. M.-J:, McManamon;P.J. and Lewis, M. G.: The pathogenesis of reflux nephropathy (chronic atrophic pyelonephritis). Brit. J. Rad., suppl., 13: 1, 1975. 3. Ransley, P. G. and Risdon, R. A.: Reflux and renal scarring. Brit. J. Rad., suppl., 14: 1, 1978. 4. Hodson, C. J. and Edwards, D.: Chronic pyelonephritis and vesicoureteric reflux. Clin. Rad., 11: 219, 1960. 5. Babcock, J. R., Keats, G. K. and King, L. R.: Renal changes after an uncomplicated antireflux operation. J. Urol., 115: 720, 1976. 6. Winberg, J., Claesson, I., Jacobson, B., Jodal, U. and Peterson, H.: Renal growth after acute pyelonephritis in childhood: an epidemiological approach. In: Reflux Nephropathy. Edited by J. Hodson and P. Kincaid-Smith. New York: Masson Publishers, pp. 309-322, 1979. 7. Claesson, I., Jacobsson, B., Jodal, U. and Winberg, J.: Early detection of nephropathy in childhood urinary tract infection. Acta Rad. (Diagn.), 22: 315, 1981. 8. Tamminen, T., Olbing, H. and Bachmann, H. J.: The development of segmental scarring in previously normal kidneys seen in three children with vesicoureteric reflux. Klin. Piidiat., 194: 137, 1982. 9. Hellstrom, M., Jacobsson, B., Jodal, U., Winberg, J. and Oden, A.:
10. 11. 12. 13. 14.
15. 16. 17.
18.
19. 20.
Renal growth after neonatal urinary tract infection. Ped. Nephrol., 1: 269, 1987. Harrison, B. R., Howards, S. S. and Thomas, R.: Medical deviation of the upper pole calix on the intravenous urogram as an indication of vesico-ureteric reflux. AJR, 126: 1189, 1976. Gedroyc, W. M. W., Chaudhuri, R. and Saxton, H. M.: Normal and near normal caliceal patterns in reflux nephropathy. Clin. Rad., 39: 615, 1988. International Reflux Study Committee: Medical versus surgical treatment of primary vesico-ureteral reflux: a prospective international study in children. J. Urol., 125: 244, 1981. Smellie, J., Edwards, D., Hunter, N., Normand, I. C. S. and Prescod, N.: Vesico-ureteric reflux and renal scarring. Kidney Int., 8: 65, 1975. Weiss, R., Tamminen-Mobius, T., Koskimies, 0., Olbing, H., Smellie, J. M., Hirche, H. and Lax-Gross, H.: Characteristics at entry of children with severe primary vesicoureteral reflux recruited for a multicente-r, international therapeutic trial eomparing medical and surgical management. J. Urol., part 2, 148: 1644, 1992. Jorulf, H., Nordmark, J. and Jonsson, A.: Kidney size in infants and children assessed by area measurement. Acta Rad. (Diagn.), 19: 154, 1978. Claesson, I., Jacobsson, B., Olsson, T. and Ringertz, H.: Assessment of renal parenchymal thickness in normal children. Acta Rad. (Diagn.), 22: 305, 1981. Jodal, U., Koskimies, 0., Hanson, E., Lohr, G., Olbing, H., Smellie, J. and Tamminen-Mobius, T.: Infection pattern in children with vesicoureteral reflux randomly allocated to operation or longterm antibacterial prophylaxis. J. Urol., part 2, 148: 1650, 1992. Hjiilmas, K., Lohr, G., Tamminen-Mobius, T., Seppanen, J., Olbing, H. and Wikstrom, S.: Surgical results in the International Reflux Study in Children (Europe). J. Urol., part 2, 148: 1657, 1992. Birmingham Reflux Study Group: Prospective trial of operative versus non-operative treatment of severe vesico-ureteric reflux: 2 years' observation in 96 children. Brit. Med. J., 287: 171, 1983. Birmingham Reflux Study Group: Prospective trial of operative versus non-operative treatment of severe vesicoureteric reflux in children: five years' observation. Brit. Med. J., 295: 237, 1987.