Renal growth after neonatal urinary tract infection

Pediatric Nephrology Pediatr Nephrol (1987) 1: 269-275 9 IPNA 1987 Original article Renal growth after neonatal urinary tract infection* Mikael Hellstr6m j, Bo Jacobsson 1, Ulf Jodal 2, Jan Winberg 3, and Anders Od6n 4 Departments of]Pediatric Radiology, 2Pediatrics, and 4Mathematical Statistics, University of G6teborg, Sweden 3Department of Pediatrics, Karolinska Hospital, Karolinska Institute, Stockholm, Sweden Abstract. This study presents the result of 12-21 years' follow-up in a group of children with neonatal urinary tract infection (onset within 1 month after birth) in whom early renal growth retardation was noted without concomitant classical renal scarring. In all cases the neonatal infection was diagnosed and treated within a few days of onset and the patients were closely supervised thereafter. Renal length, parenchymal thickness and area were measured at urography. At first follow-up (22 children, mean age 4.1 years) a significant reduction of renal parenchymal thickness was noted. Long-term follow-up (18 patients, mean age 17 years) demonstrated a normalization of renal size in the entire group, although less complete in the subgroup with reflux. There were two major findings in the present study. Firstly, renal growth retardation was seen after neonatal infection, both with and without reflux. Secondly, normalization of renal size in previously small kidneys was demonstrated, suggesting that growth retardation can be a reversible phenomenon. The tendency for such normalization was slightly more marked in children without reflux. Reduction of parenchymal thickness without calyceal deformity, therefore, does not necessarily mean irreversible damage, and differentiation between permanent scarring and temporary growth retardation can thus only be made at later follow-up, possibly not until after puberty. The demonstration of renal growth retardation in spite of early diagnosis and treatment emphasizes the great vulnerability of the kidney in the newborn. Offprint requests to: M. Hellstr6m, Department of Pediatric Radiology Childrens Hospital, East Hospital, S--41685 G6teborg, Sweden Key words: Children - Kidney - Growth retardation - Urinary tract infection - Vesico-ureteral reflux Introduction Renal damage associated with urinary tract infection (UTI) was previously named chronic atrophic pyelonephritis. Lately, the term reflux nephropathy has been introduced to emphasize the close association of renal damage to vesico-ureteral reflux (VUR) [1]. In addition to VUR, obstruction, low age, long interval between onset of infection and therapy and high virulence of the infecting bacteria have been suggested as the main risk factors for renal damage following UTI [2]. Reflux nephropathy includes several forms of renal damage. Focal scarring with narrowing of the parenchyma and clubbing of the calyces has been recognized since the initial radiological description by Hodson [3]. The renal damage may also be more diffuse, often referred to as backpressure type, similar to that seen in urinary tract obstruction [4]. However, the changes may also include retardation or cessation of renal growth [5], a phenomenon that is less well defined. Renal growth retardation has been observed in scarred as well as non-scarred refluxing kidneys [5-13]. We have previously reported the preliminary evaluation of renal growth of nonscarred kidneys after neonatal UTI with, as well as without, VUR [10]. Our findings suggested that UTI itself has an effect on the growth potential of the kidney. Orikasa et al. [8] reported that after previous growth retardation kidneys may resume normal size at cessation or improvement of VUR. 270 However, Lyon [6] considered the growth retardation to represent permanent renal damage, as in his study the kidneys never became normal in size. This study presents the results of 12-21 years' follow-up of children with neonatal UTI who developed early renal growth retardation without concomitant classical renal scarring. It also includes a re-evaluation of previous data for renal size at initial and first follow-up urography at a mean age of 4.1 years [10]. The aim of the study was to find out whether renal growth retardation may be a reversible phenomenon, and whether the outcome depends on the presence of VUR. Patients and methods Acute, neonatal, symptomatic UTI (onset within 1 month after birth) without obstruction or malformations of the urinary tract was diagnosed in 75 (54 boys, 21 girls) of approximately 57,000 infants born in the three maternity hospitals in the G6teborg area from 1960 to 1966. Of the 75 patients 69 (92%) survived the neonatal period and urography was performed in 60 of these. Follow-up urography was done in 30 children. Four patients who developed classical focal renal scarring with clubbing of the calyces a n d / o r indentation of the renal outline were excluded, as were 4 patients whose urograms were of poor quality. The remaining 22 patients (19 boys, 3 girls) were selected for the present investigation. Fifteen of these patients took part in a long-term followup investigation in 1981-1983, i.e. 15-21 years after their neonatal UTI. Three others had been subjected to urography and clinical follow-up in 1978-1979 at the ages of 12, 13 and 14 years, respectively, and were included in the study but not reexamined later. Two patients could not be traced and two declined to take part. A total of 18 patients (16 boys, 2 girls) with a mean age of 17 years (range 12-21) were thus included in the long-term follow-up, which included clinical, laboratory and radiological examinations. Recurrences of UTI had occurred in 7 patients. In total, ten episodes of asymptomatic or symptomatic recurrences were documented, all within 10 months of the initial UTI (Table 1). At long-term follow-up serum creatinine and blood pressure were within normal limits in all individuals; urinalyses demonstrated no bacteriuria, proteinuria or haematuria. None of the patients underwent urinary tract surgery during the study period. Neonatal UTI. Major presenting symptoms were fever (11 patients with a temperature of 37.9 ~ ~ weight loss or slow weight gain (7 patients), vomiting (2 patients), central nervous system symptoms (1 patient) and foul smelling urine (I patient) (Table 1). Initial UTI was diagnosed from voided urine specimens after careful cleaning of the genitals. Significant bacteriuria was defined as > 100 000 bacteria/ml urine together with a white cell count in uncentrifuged, fresh urine of 25 cells or m o r e / r a m 3 in the male and 50 or more in the female. Median white cell count in the patients included was 1300/ram3 urine (range 150-10 000/mm3). All patients were immediately treated with an appropriate antibiotic. Table 1. Clinical and radiological data on the 22 patients with neonatal urinary tract infection (UTI) Patient Sex Neonatal UTI Recurrence VUR no. 1 2 3 4 5 6 7 8 9 10 11 12 13 a 14 15 16 17 18 a 19 a 20, 21 22 M M M M M M F M M F M M F M M M M M M M M M Main presenting symptom Body temperature (~ C) Bacteria Weight Fever CNS Weight Weight Fever Fever Fever Weight Foul smelling urine Weight Vomiting Fever Fever Fever Vomiting Fever Fever Fever Fever Weight Weight 37.0 39.4 37.5 36.9 37.0 40.0 38.2 39.0 36.9 37.0 37.9 37.4 39.0 38.1 38.3 37.5 38.2 37.9 38.5 38.8 37.4 36.5 Escherichia coli E. coli E. coli E. coli E. coli E. coli Alk.Fec E. coli E. coli E. coli E. coli E. coli E. coli E. coli E. coli E. coli E. coli E. coli E. coli E. coli E. coli E. coli 3 months, 6 months 2 months, 8 months 10 months 1 month, 2 months 2 months 1 month 7 months RS LS 3 0 0 0 2 0 3 0 0 0 0 2 2 0 0 2 0 0 2 0 2 2 0 0 0 2 0 3 0 0 1 0 0 3 0 0 2 0 0 2 0 0 a No long-term follow-up M = male; F = female; Weight = weight loss or slow weight gain; CNS = central nervous system symptom; V U R = vesico-ureteral reflux, maximal observed grade (5-grade scale) on any occasion that VCU was performed; RS = right side; LS = left side 271 Further details of methods, laboratory findings and initial clinical course have been reported previously by Bergstr6m et al. [14]. Radiological evaluation. All 22 patients underwent initial urography performed after the neonatal UTI at a mean age of 33 days (range 7 - 9 9 ) and first follow-up urography at a mean age of 4.1 years (range 1-9). Long-term follow-up urography was performed in 18 patients at a mean age of 17 years (range 12-21). Measurements of kidney area were performed according to the method of Jorulf et al. [15] and of kidney length according to Ekl6f and Ringertz [16]. Renal parenchymal thickness was measured from the bottom of the minor calyx to the renal outline at three different sites (upper pole, lower pole and lateral aspect) according to the method of Claesson et al. [ 17]. Renal measurements were correlated to the L 1 - L 3 distance, i.e. the distance from the upper border of the first to the lower border of the third lumbar vertebra [16]. The results were compared with the reference material of Claesson et al. [17], supplemented with new reference material for ages 16-25 (Claesson, unpublished data). The latter was obtained in a way similar to that described by Claesson et al. [17]. In addition, a careful evaluation of kidney parenchyma and collecting system was carried out by two radiologists (M. H. and B. J.). Nineteen children underwent voiding cystourethrography (VCU) at a mean age of 39 days (range 8-78), while 2 others had their first VCU at 7 months and 3.3 years, respectively. One patient was never examined by VCU. A second VCU was performed in 15 children at a mean age of 2.2 years (0.25-4.6 years). Long-term follow-up VCU was performed in 5 patients at a mean age of 14.1 years (range 9.8-17.6). VUR was graded I - V according to the IRSC (International Reflux Study in Children) grading system [18]. Kidneys with V U R on any occasion that VCU was performed were included in the VUR group (Table 1). Statistical methods. Comparisons between the present material and the reference material were performed using raw data from individuals in the reference materials with L 1 - L 3 distances similar to those in the present series. All evaluations were based on comparisons of groups of children. The mean of the right and left kidney measurements was used in the statistical evaluation. The patients were subdivided into a refluxing group (uni- or bilateral VUR) and a non-refluxing group (bilateral absence of VUR). To avoid the possible effects of contralateral compensatory hypertrophy in patients with unilateral VUR, only the measurements on the refluxing side were used for evaluation. The possibility of unilateral renal growth retardation with development of contralateral compensatory hypertrophy was tested by comparing renal size asymmetry at first and longterm follow-up, using the quantity [ ( r i g h t - left)/(right + left) [ to define the degree of asymmetry between parenchymal thickness measurements on both sides. In order to eliminate the influence of the L 1 - L 3 distance, a special test technique suggested by Mantel [19] was used for all comparisons between groups. The p-values given refer to two-sided tests. A p-value of < 0.05 was considered statistically significant. Results Of the 21 neonates investigated with initial VCU after the neonatal UTI, 9 had VUR (1 renal unit with grade I, 11 with grade II and 3 with grade III). At first follow-up, VUR was found in 3 of 15 children (1 renal unit with grade I, 2 with grade II and 1 with grade III). A long-term follow-up VCU was performed in 5 patients, 4 of whom had VUR at previous VCUs. VUR was demonstrated in 2 patients (1 renal unit with grade I and 2 with grade II). In total, VUR was demonstrated at some occasion in 9 patients, 7 of whom had long-term follow-up urography. The maximal observed grade of VUR is demonstrated in Table 1. Of the 12 patients who had no demonstrable VUR, 10 had long-term follow-up urography. Urography demonstrated no signs of calyceal deformities indicative of obstruction or classical renal scarring at any of the examinations. Measurements of renal length, parenchymal thickness and area as compared with the reference material are given in Table 2. At initial urography only 44 of 110 possible renal measurements (5 measurements in each of the 22 patients) were obtainable, mainly due to bowel gas obscuring the renal outlines (15 patients contributed with one, two or more renal measurements, while in 7 patients no measurements could be obtained). Comparison of these 44 measurements with the reference material showed that parenchymal thickness was slightly smaller in these patients, but the differences were not statistically significant (Table 2; Figs. 1, 2). At first follow-up (22 patients) 109 of 110 possible renal measurements were obtainable, and at long-term follow-up (18 patients) 89 of 90 possiTable 2. Renal length, parenchymal thickness (upper pole, lower pole and lateral aspect) and area at initial and follow-up urography (at a mean age of 4.1 and 17.0 years, respectively) compared with reference material ([17]; Claesson, unpublished data) with corresponding L 1 - L 3 distance (Mantel's test, two-sided) Initial urography Length Upper pole Lower pole Lateral aspect Area (n = 15) a p-value First follow-up urography (n = 22) p-value Long-term follow-up urography (n = 18) p-value 0.32 0.06 0.06 0.05 0.31 0.30 < 0.01 <0.001 < 0.01 0.18 0.27 0.38 0.13 0.46 0.46 At initial and first follow-up urography, were smaller in the present material than erial a Fifteen patients contributed with one, measurements, while no measurements seven patients all renal parameters in the reference mattwo or several renal were obtainable in 272 Lower (cm) pole II I n=13 4.5. n=22 III ', I n=18 :~mean 4.0- + 2SD i 3.53.0- 9 ___.-2SD 2.52.01.51.0- 9 i 0.5O" " - ~ , , , , , , , i , ~ l 3 4 5 6 7 8 9 10 11 12 13 ~ (cm) L 1-L j +2SD 3 Fig. 1. Renal growth after neonatal urinary tract infection (UTI). Parenchymal thickness measured at the lower kidney pole, correlated to the length of a lumbar segment of the spine ( L I - L 3 distance). Comparison with reference material ([17]; Claesson, unpublished data). I = initial urography; II = first follow-up urography (mean age 4.1 years); III = long-term follow-up urography (mean age 17 years), n = number of patients A r e a (cm 2) J I n=6 80- II n=22 III n=18 mean 70 ~ 60- ~ ". " ~ -2SD 5040302010O' -% 0 , , ~ ~ ~ ~ 3 4 5 6 7 8 , g ~ 10 J 11 Table 3. Renal length, parenchymal thickness (upper pole, lower pole and lateral aspect) and area at first follow-up urography (22 patients, mean age 4.1 years) compared with reference material ([17]; Claesson, unpublished data) with corresponding L1 - L 3 distance (Mantel's test, two-sided) LI - L 3 (cm) Lenght (cm) Upper pole (cm) Lower pole (cm) Lateral aspect (cm) Area (cm 2) Present series (mean _+ SD) Reference material (mean) ~ 6.24+0.79 8.60+0.91 2.18+0.29 2.07 + 0.24 1.60+0.19 25.20+4.98 6.24 8.78 2.41 2.28 1.75 26.50 , 12 ~ 13 ~- L 1 - L 3 (cm) Fig. 2. Renal growth after neonatal UT1. Renal area correlated to the length of a lumbar segment of the spine (L1-L3 distance). Comparison with reference material ([17]; Claesson, unpublished data). I = initial urography; II = first follow-up urography; III = long-term follow-up urography; n = number of patients Table 4. Renal length, parenchymal thickness (upper pole, lower pole and lateral aspect) and area at long-term follow-up urography (18 patients, mean age 17 years) compared with reference material ([17]; Claesson, unpublished data) with corresponding LI - L 3 distance (Mantel's test, two-sided) p-value 0.30 <0.01 < 0.001 <0.01 0.18 L 1 - L 3 (cm) Lenght (cm) Upper pole (cm) Lower pole (cm) Lateral aspect (cm) Area (cm 2) Present series (mean _+ SD) Reference material (mean)" 11.36+1.11 13.36 + 1.11 3.29 + 0.39 3.14__.0.35 2.40 _ 0.36 58.43 _+8.26 11.36 13.24 3.38 3.28 2.48 58.09 p-value 0.27 0.38 0.13 0.46 0.46 " Estimated mean of renal measurements - determined by regression analysis - with mean and range of L1 - L3 distance in the reference material equal to that of the present series Estimated mean of renal measurements - determined by regression analysis - with mean and range of L1 - L3 distance in the reference material equal to that of the present series ble measurements were made. At first follow-up urography a significant reduction of parenchymal thickness was demonstrated (Tables 2, 3; Figs. 1, 2). Renal length and area were also slightly smaller, but the differences did not reach statistical significance (Tables 2, 3). At long-term follow-up, the renal parameters for the group were well within normal limits, indicating catch-up of the previ- ously subnormal parenchymal thickness (Tables 2, 4; Figs. 1, 2). Separate analysis of the group with and the group without VUR revealed that all renal parameters in both groups were smaller than in the reference material at first follow-up (Table 5). At long-term follow-up all renal parameters in the non-refluxing group were comparable with the 273 Table 5. Renal length, parenchymal thickness (upper pole, lower pole and lateral aspect) and area at first and long-term follow-up urography in the groups with and without VUR compared with reference material ([17]; Claesson, unpublished data) with corresponding L 1 - L 3 distance (Mantel's test, two-sided) Length Upper pole Lower pole Lateral aspect Area First follow-up ~ Long-term follow-up a VUR (n = 9) p-value No VUR (n = 12) p-value VUR (n = 7) p-value No VUR (n = 10) p-value 0.25 <0.05 < 0.001 <0.01 0.06 0.19 <0.01 0.07 0.06 0.21 0.90 0.08 < 0.05 0.30 0.50 0.40 0.70 0.69 0.98 0.36 VCU was not performed in one patient reference material, indicating full normalization of renal size (Table 5). In the group with VUR, the lower pole was slightly reduced while the other parameters were within normal limits, indicating nearly complete normalization of renal size (Table 5). Comparison at first follow-up urography between the groups with and without VUR demonstrated that the VUR group had smaller renal measurements than the non-refluxing group, but the differences were not statistically significant. At long-term follow-up the group with VUR (grade I - I I I ) had significantly smaller parenchymal thickness at the upper and lower poles as compared with the non-refluxing group (p< 0.05) (data not shown). The degree of asymmetry between right and left kidney size was similar at first and long-term follow-up (p=0.40-0.91), indicating that compensatory hypertrophy (in the case of unilaterally small kidneys) did not explain the normalization of mean renal size noted at long-term follow-up. Recurrent UTI occurred in seven patients, in all cases before 10 months of age. Renal parameters in this group were not different from those in the group without UTI recurrences. Discussion Infancy, especially the neonatal period, appears to be the most vulnerable age for developing renal damage after UTI. It is also the age at which the rate of growth of the kidney is at its highest [5]. Therefore retardation of growth should be easier to detect during this period than later in life. It thus seems logical to select neonates for studying renal growth after UTI. Acute infection of the growing kidney may cause different types of damage, focal scarring and renal growth retardation, which may occur together or separately. Their pathogenesis might be different [20]. Most studies on renal growth retardation have included scarred kidneys, where it is difficult to differentiate subnormal growth from progression of existing scars or extension of scarring to previously unscarred parts of the kidney. Also, compensatory hypertrophy of non-damaged parts may make evaluation difficult. Thus, the phenomenon of growth retardation as such is best studied on non-scarred kidneys. Accelerated renal growth at puberty has been demonstrated in scarred kidneys with growth retardation after UTI [21]. Long-term follow-up therefore seems important for the final evaluation of renal growth disturbances. In previous studies renal growth has usually been assessed by measurements of renal length. Employing such measurements or renal area measurements in the present study would not have revealed the subnormal kidney size that was obvious only by measurements of parenchymal thickness (upper pole, lower pole and lateral aspect). It has been demonstrated by Claesson et al. [22] that measuring renal parenchymal thickness is more sensitive than length and area measurements in detecting subtle deviations from normal kidney size. The importance of carefully performed parenchymal measurements for the early detection of renal damage was also emphasized by Hodson [5]. The statistical evaluations in the present study were based on non-parametric comparisons with raw data from the reference material ([17]; Claesson, unpublished data). In the previous preliminary statistical evaluation [10], as in most similar analyses by others, raw data from reference materials have not been used. Using estimated regression parameters instead of raw data for statistical comparisons means a loss in accuracy, which should be avoided, if possible. There were two major findings in the present study. Firstly, renal growth retardation was seen after neonatal UTI, both with and without VUR. Secondly, normalization of renal size in previously small kidneys was demonstrated, suggesting that growth retardation can be a reversible phenomenon. The tendency for such normalization was slightly more marked in children without VUR. The present study demonstrates the extreme vulnerability of the neonatal kidney and at the same time its capacity to resume growth under favourable conditions. Thus, the 44 kidneys in- 274 cluded in this study showed - as a group - significant growth retardation that was demonstrable a mean of 4 years after the acute infection. This impairment of growth potential occurred although in most instances the infections were diagnosed and treated immediately after the onset of symptoms. We can only speculate that the catchup growth demonstrated at a mean of 17 years after infection was dependent on the early therapy these children received. With a delay of therapy the outcome might have been different. It should also be recalled that in the entire group of 69 surviving children given early therapy, focal renal scars were demonstrated in only 4 cases (not included among the 22 children studied in this paper). At initial urography there was a statistically non-significant tendency towards reduced renal parenchymal thickness. Since only 40% of the renal measurements were obtainable, the figures must be judged with caution. Thus, the diagnosis of growth retardation,as compared with the reference material, was not based on the findings at initial urography, but on first follow-up urography, where 109 of 110 renal measurements were obtainable. However, renal growth is normally very rapid in the neonatal period, and the interval between onset of infection and initial urography was, at least in some patients, long enough to allow early renal growth retardation to be detected. The possibility of congenitally small kidneys cannot be excluded, but appears less likely considering the absence of other morphological abnormalities and the full normalization of kidney size that occurred later. In most children the renal growth disturbance was probably initiated by one single neonatal infection, or by recurrent infection occurring before 10 months of age (7 patients). The results of experimental studies suggest that cell multiplication in the kidney is at a maximum in the early postnatal period [23, 24], making the kidney especially vulnerable. Recent experimental data also suggest that cell multiplication continues until adulthood, although at a much lower rate [24, 25], which would explain the capacity of catch-up growth in non-scarred kidneys. Our findings would be compatible with the hypothesis that infection not leading to extensive cell destruction may interfere transiently with the cell multiplication process. All the mechanisms by which renal growth retardation occurs are not known. Growth retardation after obstruction [4] or gross VUR without urinary tract infection has been attributed to hy- drodynamic factors and alterations in renal blood flow [8]. In the case of infection without obstruction or VUR other factors seem to be involved. Asscher et al. [20] demonstrated that live as well as heat-killed bacteria caused renal growth impairment in rats, while only live bacteria caused scarring. The authors concluded that scarring and growth retardation do not share a common pathogenesis and suggested that bacterial endotoxin might be the factor responsible for such growth impairment. Orikasa et al. [8] studied the growth rate of kidneys in children with VUR; they stated that only few growth retarded kidneys returned to normal size after cessation or improvement of VUR, while most remained subnormal in size 1- 8 years later. The phenomenon of accelerated (faster than normal) renal growth rate has been described soon after surgical correction of VUR [26, 27] and also in scarred kidneys at the time of puberty [21]. Lyon [6] described continued renal growth after temporary growth arrest, but pointed out that the kidneys never reached normal size after a mean follow-up of 4.5 years (range 1-15). The discrepancy between our results and those of Lyon can be explained by the longer follow-up time in the present study, allowing for later growth spurt, perhaps at puberty. It can also be assumed that the potential for normalization of kidney size is less in kidneys with advanced growth retardation, as in Lyon's study. Renal growth retardation in our study was only moderate, thus possibly allowing for more complete catch-up growth. The finding at long-term follow-up that kidneys with VUR, albeit catching up to normal size, were slightly smaller than kidneys without VUR may perhaps be explained by larger amounts of bacteria reaching the kidney subject to VUR, thus inducing a more severe inflammatory response [28]. The possibility that VUR as such induced the growth retardation appears less likely, especially as VUR was of only moderate grade in the majority of cases. In conclusion, it has been demonstrated that renal growth retardation may occur after neonatal UTI, not only with, but also without VUR. It also appears that renal size may normalize after previous growth retardation. The present study was highly selective, and the results are therefore not directly applicable to other groups of children with UTI. However, the possibility of renal size normalization should be taken into account in the follow-up of growth-retarded kidneys. Thus, reduction of parenchymal thickness in the absence of calyceal deformity or 275 indentation of the renal outline does not necessarily mean irreversible damage to the kidney. The definite differentiation between a permanent scar and temporary growth retardation can, at urography, only be made at later follow-up, possibly not until after puberty. The demonstration of renal growth retardation in spite of early diagnosis and treatment emphasizes the great vulnerability of the kidney in the newborn. Acknowledgements. This investigation was supported by grants from The Swedish Society of Medical Radiology, the G6teborg Medical Society, Agfa-Gevaert AB, Sweden and the Swedish Medical Research Council, grant no. 765. Statistics were performed by Anders Od6n, Ph.D. References 1. Bailey R (1973) The relationship of vesico-ureteric reflux to urinary tract infection and chronic pyelonephritis - reflux nephropathy. Clin Nephrol 1: 132-141 2. Winberg J, Bollgren I, Kfillenius G, M611by R, Svensson SB (1982) Clinical pyelonephritis and focal renal scarring. Pediatr Clin North Am 29:801-814 3. Hodson CJ (1959) The radiological diagnosis of pyelonephritis. Proc R Soc Med 52:669-672 4. Hodson CJ, Craven JD (1966) The radiology of obstructive atrophy of the kidney. Clin Radiol 17:305-320 5. Hodson CJ, Davies Z, Prescod A (1975) Renal parenchyrnal radiographic measurement in infants and children. Pediatr Radiol 3:16-19 6. Lyon RP (1973) Renal arrest. J Urol 109:707-710 7. Redman JF, Scriber LJ, Bissada NK (1974) Apparent failure of renal growth secondary to vesicoureteral reflux. Urology 3:704-707 8. Orikasa S, Takamura T, Inada F, Tsuji I (1978) Effect of vesicoureteral reflux on renal growth. J Urol 119:25-30 9. Wikstad I, Aperia A, Broberger O, Ekengren K (1979) Vesicoureteric reflux and pyelonephritis. Long time effect on area of renal parenchyma. Acta Radiol (Diagn) 20: 252-260 10. Winberg J, Claesson I, Jacobsson B, Jodal U, Peterson H (1979) Renal growth after acute pyelonephritis in childhood: an epidemiological approach. In: Hodson J, Kincaid-Smith P (eds) Reflux nephropathy. Masson, New York, pp 309-322 11. Smellie JM, Edwards O, Normand ICS, Prescod N (1981) Effect of vesicoureteric reflux on renal growth in children with urinary tract infection. Arch Dis Child 56:593-600 12. Walker RD, Richard GA, Fennell RS, Iravani A, Garin E (1983) Renal growth and scarring in kidneys with reflux and a concentrating defect. J Urol 129:784-786 13. Ginalski J-M, Michaud A, Genton N (1985) Renal growth retardation in children: sign suggestive of vesicoureterat reflux? AJR 145:617-619 14. Bergstr6m T, Larsson H, Lincoln K, Winberg J (1972) Studies of urinary tract infections in infancy and childhood. XII. Eighty consecutive patients with neonatal infection. J Pediatr 80:858-866 15. Jorulf H, Nordmark J, Jonsson A (1978) Kidney size in infants and children assessed by area measurement. Acta Radiol (Diagn) 19:154-62 16. Ekl6f O, Ringertz H (1976) Kidney size in children. A method of assessment. Acta Radiol (Diagn) 17:617-625 17. Claesson I, Jacobsson B, Olsson T, Ringertz H (1981) Assessment of renal parenchymal thickness in normal children. Acta Radiol (Diagn) 22:305-314 18. IRSC, International Reflux Study in Children (1985) International system of radiographic grading of vesicoureteric reflux. Pediatr Radiol 15:105-109 19. Mantel N (1963) Chi-square tests with one degree of freedom; extensions of the Mantel-Haenszel procedure. J Am Statist Assoc 58:690-700 20. Asscher AW, Chick EM, Evans KT, Williams JE (1971) Effect of ascending Escherichia coli infection on compensatory hypertrophy of the rat kidney. In: Kincaid-Smith P, Fairley KF (eds) Renal infection and renal scarring. Mercedes, Melbourne, pp 303-313 21 Claesson I, Jacobsson B, Jodal U, Winberg J (1981) Compensatory kidney growth in children with urinary tract infection and unilateral renal scarring: an epidemiologic study. Kidney Int 20:759-764 22. Claesson I, Jacobsson B, Jodal U, Winberg J (1981) Early detection of nephropathy in childhood urinary tract infection. Acta Radiol (Diagn) 22:315-320 23. Winick M, Noble A (1965) Quantitative changes in DNA, RNA and protein during prenatal and postnatal growth in the rat. Dev Biol 12:451-466 24. Celsi G, Jakobsson B, Aperia A (1987) Influence of age on compensatory renal growth in rats. Pediatr Res (in press) 25. Sands J, Dobbing J, Gratrix CA (1979) Cell number and cell size: organ growth and development and the control of catch-up growth in rats. Lancet II: 503-505 26. Atwell JD, Vijay MR (1970) Renal growth following reimplantation of the ureters for reflux. Br J Urol 50:367-370 27. Willscher MK, Bauer SB, Zammuto PJ, Retik AB (1976) Renal growth and urinary infection following antireflux surgery in infants and children. J Urol 115:722-725 28. Bille J, Glauser MP (1982) Protection against chronic pyelonephritis in rats by suppression of acute suppuration: effect of cholchicine and neutropenia. J Inf Dis 146: 220-226 Received March 18, 1986; received in revised form October 10, 1986; accepted February 11, 1987