- Email: [email protected]
Dysplastic kidneys in children – Do they grow?
Journal of Pediatric Urology (2010) 6, 40e44
Dysplastic kidneys in children e Do they grow? Nia Fraser a,*, Anu Paul a, A.R. Williams a, N. Broderick b, M.U. Shenoy a a
Department of Paediatric Urology, Nottingham University Hospitals NHS Trust, Queens Medical Centre Campus, Level E, East Block, Derby Road, Nottingham NG7 2UH, UK b Department of Paediatric Radiology, Nottingham University Hospitals NHS Trust, Queens Medical Centre Campus, Level E, East Block, Derby Road, Nottingham NG7 2UH, UK Received 4 February 2009; accepted 11 May 2009 Available online 30 June 2009
KEYWORDS Dysplastic kidney; PUV; VUR; Ultrasound
Abstract Objectives: Dysplastic kidneys (DK) are a common cause of chronic kidney disease (CKD). Little is known about their growth or how their sonographic appearance changes. This study aimed to test the hypothesis that DK gain little length, and to identify radiologic trends predictive of CKD. Methods: Ultrasound scans of children with DK born in 1980e2005 and referred to a single tertiary centre were analysed by a pediatric radiologist. Renal lengths were plotted on standard nomograms and the degree of dysplastic appearance noted. Factors related to DK e bladder outlet obstruction, vesico-ureteric reflux and renal impairment e were noted. Results: Fifty-three children were studied (83 kidneys), of whom 41 were boys; 289 scans were analysed. In 33 children there was associated bladder outlet obstruction or vesico-ureteric reflux. Forty-four DK were noted to fall off their renal length ‘centile’. This correlated well with the development of CKD and is statistically significant. Deterioration occurred in 53% of DK; primarily progressive reduction in corticomedullary differentiation. This also correlated well with development of CKD. Conclusion: More than half of the DK showed poor growth velocity. This, together with the degree of sonographic abnormality, carries a high predictive value for development of CKD. We recommend diligent serial sonography to follow renal growth and dysplastic appearance in children with DK. ª 2009 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved.
Introduction Dysplastic kidneys (DK) are an important cause of chronic kidney disease (CKD) in children [1]. They may arise de novo
* Corresponding author. Tel.: þ44 115 924 9924x62626. E-mail address: [email protected] (N. Fraser).
(primary) or be found in association with congenital obstruction of the urinary tract or vesico-ureteric reflux (VUR) (secondary). Certain genes, transcription and growth factors have been implicated in the development of DK but there is a paucity of information in the medical literature regarding the growth potential of these abnormal kidneys. This study set out firstly to test the hypothesis that DKs do not grow in length significantly, as measured by serial
1477-5131/$36 ª 2009 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jpurol.2009.05.014
Dysplastic kidneys in children postnatal ultrasonography. Secondly, we aimed to identify any sonographic features which were predictive for chronic renal failure.
41 interpretation of US (such as poor quality images, and some ectopic kidneys which were difficult to measure accurately).
Antenatal findings
Methods Case notes of all children with DK born between 1980 and 2005 referred to a single tertiary pediatric nephrourology centre were analysed. Sonographic criteria for the diagnosis of renal dysplasia were combinations of the presence of small-to-normal-sized kidneys showing echogenic parenchyma, reduction or loss of corticomedullary differentiation, and cortical cysts (Fig. 1). Postnatal ultrasound (US) studies were performed using a 5- and 7-MHz variable-focus transducer. All available renal tract US films were reviewed by a single senior radiologist in one institution. Longitudinal renal lengths were plotted against length or height of child on standard European nomograms (Fig. 2), the rationale being that published charts show an approximately linear relation between renal length and height in children over 1 year old [2]. This also avoids the additional variability of height with age, particularly important in children in our study population who may have CKD-related reduction in height velocity. In a very small number of cases film copies were unavailable, and reports of the US scans performed by the same radiologist or his colleagues were used. Children who had not undergone follow-up US were excluded, as were those whose scans were of poor image quality, as judged by the radiologist. Details drawn from the case notes included antenatal US findings, age at and mode of presentation, initial creatinine, and associated congenital abnormalities. The number of children with primary DK and those secondary to underlying pathology were established. The outcomes were plots of serial renal lengths recorded on nomograms compared to expected renal growth, and changes in dysplastic appearance. Also noted were survival, development of CKD and end-stage renal failure (ESRF). Assessment of degree of CKD was based on estimated GFR with reference to the Schwartz formula in most children, unless due to severe failure to thrive and poor body mass the creatinine did not reflect the child’s renal function. In these children, formal EDTA GFR was performed to attain a more accurate reflection of renal function. CKD was defined according to the National Kidney Foundation guidelines [3] (stages 1e5: stage 1, normal GFR but evidence of renal damage, stage 2 GFR 60e89, stage 3 GFR 30e59, stage 4 GFR 15e29, and stage 5 ESRF with GFR < 15 and on dialysis). For each renal US reviewed, the following parameters were recorded as well as renal length: age at first postnatal US, interval between subsequent scans, whether the DK were unilateral or bilateral, appearance of cortex (echogenicity, corticomedullary differentiation, presence of cysts, cortical loss or scarring), remainder of the urinary tract.
Results A total of 53 children were included in the study (83 renal units). Forty one (77%) of these were boys. Ten children were excluded due to incomplete data, i.e. difficulty in
In 32 children (60%), abnormal renal tracts were identified prenatally. Nine children had apparently normal renal tracts and a further three mothers were reported to have had a normal pregnancy. Results of fetometernal scan (FMS) were unrecorded in nine children. The majority of antenatal findings were first noted during the 20/40 fetal anomaly US, but some were picked up at 16/40, 18/40, 30/40 and 34/40. The commonest abnormalities were hydronephrosis (53%), small or discrepant renal size (34%) and oligohydramnios (34%). Other findings included echogenic cortex, cysts and ureteric dilation. PUV were suspected in seven boys. Thirty children with renal abnormalities on FMS underwent their first postnatal US at a mean of 35 days (range 1 daye8.5 months). The other two children presented later with symptoms. Both, in retrospect, had abnormalities on their antenatal scan. In the first child, antenatal bilateral hydronephrosis was seen but he was not referred until he developed a UTI at 5 weeks of age. In the second child, oligohydramnios was found during early FMS but this had settled during the pregnancy. He was not investigated until he presented 5 years later with enuresis.
Other modes of presentation Eleven out of the remaining 21 children presented during the neonatal period: one with UTI at 2 weeks of age, five with raised creatinine (two picked up on routine neonatal blood tests in premature babies), and three were screened due to other abnormalities (VACTERL, lower mesodermal defect sequence, two-vessel umbilical cord). Two boys were noted to have a palpable bladder and poor urinary stream during the first week of life. One of these was subsequently found to have PUV and the other neuropathic bladder. The remaining 10 children presented later with the following: UTI (6, at ages 5 weeks, 2 months, 5 months, 6 months, 15 months and 6 years), high creatinine (1), renal tract screening due to history of familial renal dysplasia (1, age 2 years), screening due to two-vessel umbilical cord at age 2 months (1), investigation for short stature at 4 years of age.
Primary and secondary DK The number of children with primary DK (defined as no evidence of underlying factors such as VUR, PUJ obstruction or bladder outlet obstruction) was nine. There were 33 children in whom an underlying cause could be identified, based on results of VCUG, postnatal US and other investigations. The commonest underlying factor was VUR, followed by PUV, PUJ obstruction and one each of bladder agenesis, neuropathic bladder and congenital urethral stricture. In the remaining 11 children with DK, it was not possible to clarify whether these were primary or secondary (Table 1).
Figure 1 Images 1 and 2: supine and prone scans of a normal right kidney. Renal cortex is less echogenic than adjacent liver and there is clear differentiation of cortex from pyramids in the medulla and central calyceal echoes. Image 3: small (3 cm) dysplastic right kidney on prone scan with echogenic cortex but some preservation of corticomedullary differentiation (CMD). Images 4 and 5: supine and prone scans of a good-sized dysplastic right kidney with echogenic cortex and loss of CMD, but no cysts. Image 6: small (2.5 cm) dysplastic right kidney on supine scan with echogenic cortex and loss of CMD, but no cysts. Image 7: small (3 cm) dysplastic right kidney on supine scan with echogenic cortex, loss of CMD, and several small cortical cysts. Image 8: two prone images of a good size (6.6 cm) dysplastic right kidney with echogenic cortex, loss of CMD, and multiple cysts e three in peripheral location posteriorly at the upper pole (right hand image), one in peripheral location anteriorly at the lower pole, and one deeper lower pole cortical cyst (left-hand image). Largest cyst, 6 mm maximum diameter.
Dysplastic kidneys in children
43 performed per child was three (range 2e6) and the mean interval between scans was 2 years (range 2 monthse6 years). Nineteen US scans were not available for analysis (reports used instead). Two were ectopic but measurements were possible. When the plots were analysed, 44 (53%) DK were noted to fall off their initial renal length ‘centile’, or fell further away from the -2 standard deviation (2 SD) line in the case of the very small DK. Ten of these kidneys could not be visualized on follow-up scans and were presumed to be very small. Nineteen grew along their initial plotted line. Twenty kidneys grew sufficiently to gain SDs.
Appearance of DK on initial US
Twenty-four children (44%) had at least one other congenital abnormality. These included syndromes and associations (5), other defects in the genitourinary tract, and miscellaneous anomalies. Twelve children with DK were born prematurely (20%), between 30/40 and 37/40 gestation [30/40 (1), 32/40 (3), 33/40 (2), 34/40 (1), 35/40 (1), 36/40 (3), 37/40 (1)]. IUGR was present in three children.
Four kidneys had so little cortex that the presence of increased echogenicity, reduced corticomedullary differentiation (CMD) and cysts was not possible to ascertain. The commonest sonographic feature indicative of renal dysplasia in the population studied was increased echogenicity (91%), followed by absent or reduced CMD (86%), and cortical cysts (30%). Worsening appearance was manifested by increasingly echogenic cortex, progressive reduction in CMD and appearance of new cysts. Deterioration (seen in 53%), and most commonly progressive reduction in CMD, correlated well with the development of renal insufficiency. This was statistically significant (95% CI of 35e71%). Conversely, of children maintaining normal renal function, most (92%) had either no change or improvement in renal appearance.
Laterality
Survival
Bilateral DK (32 children in all) were commoner than unilateral DK, the majority in boys. Right was slightly commoner than left; 44 and 39, respectively. In the nine children with primary DK, four were bilateral (44%). Of the children with secondary DK 70% were bilateral.
At the time of writing there were 52 survivors. One child with ESRF and Allagille syndrome died at the age of 13.5 months, but not as a direct complication of renal dysplasia.
Figure 2 Nomogram of renal length versus child’s height [2]. Red line demonstrates a kidney falling off its growth centile. Used with permission granted by Elsevier and Pediatric Radiology.
Congenital abnormalities and prematurity
CKD
Renal length in DK A total of 289 renal images were analysed and renal lengths measured and plotted on standard nomograms of renal length versus child’s height (Fig. 2). The mean number of US scans Table 1 Underlying (‘secondary’).
renal
tract
pathology
in
DK
Unilateral Bilateral Primary (n Z 9) 5 Secondary (n Z 33) VUR 8 PUV 0 PUJ obstruction 2 Neuropathic 0 bladder Urethral stricture 0 Bladder agenesis 0 Not ascertained 6 (n Z 11) Total: 53 21
4 11 8 1 1 1 1 5 32
The serum creatinine was normal throughout in a total of 13 children. Forty children (76%) had raised creatinine on presentation or shortly afterwards (above 120 mmol/L in 24) and in eight of these the creatinine normalized within 2 years. The remaining 31 surviving children went on to develop CKD. Three had EDTA GFR, the rest estimated GFR. Sixteen children developed mild renal impairment not requiring dialysis. The remaining 15 children have CKD or ESRF. Three children are currently receiving peritoneal dialysis or haemodialysis. Nine children have undergone renal transplantation, median age 6.4 years, seven of which were pre-emptive. Three children are awaiting transplant. One child’s transplant was complicated by venous thrombosis and he subsequently received a further deceased donor transplant. In children with CKD or ESRF four main patterns were noted: (1) bilateral DK: either both well below the 2 SD below mean line throughout, or one or both dropping in SD (commonest pattern, 34%) (Fig. 2);
44 (2) bilateral DK with good growth, no scarring but with underlying VUR or bladder outlet obstruction (28%); (3) solitary DK (19%): due to contralateral pathology e renal agenesis, multicystic dysplastic kidney, nephrectomy; half of these were growing, the other half had poor growth; (4) bilateral DK with good growth but global cortical loss or scarring (16%). Only one child with disappearing DK and a normal contralateral kidney developed chronic renal failure. All who have undergone or are awaiting renal transplantation had either bilateral DK with poor growth, solitary DK or scarred DK. Apart from renal impairment and its consequences, UTIs were a common finding (20 children) and were recurrent in 14 children.
Follow up Median duration of follow up was 7 years (range 1e18 years). All children with renal impairment have ongoing follow up with nephrologists. Ten with normal renal function were discharged and two lost to follow up.
N. Fraser et al. of the scans) review the US results. We accept that measuring both renal lengths sonographically and heights/ lengths in children can be difficult and may be a cause of inaccurate plotting on the nomograms. We recognise that over the time period US technology has improved; this is less likely to have impacted on renal measurements than renal appearance (the more subtle abnormalities may have been more readily detected in the latter years studied). There was a tendency to reject more of the earlier US results, where measurements were inaccurate or could not be estimated (usually US performed elsewhere, where low frequency probes were used, limiting spatial resolution), thus strengthening our data. We acknowledge that due to a relatively small number of US scans per patient our renal growth plots may merely indicate a trend. However, previous studies have shown that the observer variation in the measurement of renal length and size is equivalent to 2e3 years’ normal growth [4]. Therefore, more frequent scanning would not necessarily have been helpful over the time period studied. A longer duration, preferably prospective, study would be useful to confirm our findings.
Conclusion Discussion Dysplastic kidneys account for up to 14% of cases of CKD [1] and occur as a result of failure of normal fetal renal maturation. It is generally accepted that the diagnosis is reached sonographically, typified by small to normal sized, echogenic kidneys with reduction or loss of CMD, with or without cortical cysts. Histologically, dysplasia is characterized by the existence of primitive mesenchymal structures, cortical cysts and the primitive duct, the product of the ureteral bud system. Numerous studies have investigated the mechanisms involved in the development of renal dysplasia: the role of fetal urinary tract obstruction [5e7], the WNT-1 gene, PAX2, BCL2, galectin 3, transforming growth factor B1, and various types of transcription factors, cytokines and cell adhesion molecules [8,9]. To our knowledge, this is the first study which has objectively assessed renal growth in DK, correlating poor growth with the development of CKD. Furthermore, we have correlated worsening sonographic appearance with CKD. Overall, 53% of DK fell short of expected growth when plotted on a nomogram of renal length versus child’s height, despite some of them achieving an increase in length. This correlates well with the development of CKD and is statistically significant (44 out of 83 DK, 95% CI 42e64%). Solitary, poorly growing DK; bilateral, tiny DK/those dropping below the line of expected growth; scarred DK; and interestingly a group with bilaterally well growing DK with underlying PUV or VUR, were the main culprits for poor renal outcome in our series. More than half of the children going on to develop CKD independently had worsening dysplastic changes visible on serial US, which was also statistically significant. We eliminated inter-observer variation by having one experienced radiologist (who had himself performed most
As a result of our findings, we recommend that renal lengths of children with DK should be recorded on a standard nomogram, and a careful search attempted to pick up the subtle signs of worsening dysplastic appearance. Knowing that a fall-off in growth and worsening appearance both significantly raise the chances of developing CKD is important, and should prompt the clinician to ensure that close nephrological input is commenced and maintained.
References [1] Pistor K, Scharrer K, Olbing H, Tamminen-Mobius T. Children with chronic renal failure in the Federal Republic of Germany: primary renal diseases, age and intervals from early renal failure to renal death. Clin Nephrol 1985;23:278e84. [2] Dinkel E, Ertel M, Dittrich M, Peters H, Berres M, Sculte-Wissermann H. Kidney size in childhood. Sonographical growth charts for kidney length and volume. Pediatr Radiol 1985;15:38e43. [3] National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification and stratification. Am J Kidney Dis 2002;39(1):S1e266. [4] Sargent MA, Wilson BPM. Observer variability in the sonographic measurement of renal length in childhood. Clin Radiol 1992;46: 344e7. [5] Hoover DL, Duckett JW. Posterior urethral valves, unilateral reflux and renal dysplasia: a syndrome. J Urol 1982;128:994e7. [6] Shibata S, Shigeta M, Shu Y, Watanabe T, Nagata M. Initial pathological events in renal dysplasia with urinary tract obstruction in utero. Virchows Arch 2001;439:560e70. [7] Spencer JR, Maizels M. Inhibition of protein glycosylation causes renal dysplasia in the chick embryo. J Urol 1987;138:984e90. [8] Glassberg KI. Normal and abnormal development of the kidney: a clinician’s interpretation of current knowledge. J Urol 2002; 167:2339e51. [9] Yang SP, Woolf AS, Yuan HT, Scott RJ, Risdon RA, O’Hare MJ, et al. Potential biological role of transforming growth factor-B1 in human congenital kidney malformations. Am J Pathol 2000; 157(5):1633e47.
Dysplastic kidneys in children e Do they grow? Nia Fraser a,*, Anu Paul a, A.R. Williams a, N. Broderick b, M.U. Shenoy a a
Department of Paediatric Urology, Nottingham University Hospitals NHS Trust, Queens Medical Centre Campus, Level E, East Block, Derby Road, Nottingham NG7 2UH, UK b Department of Paediatric Radiology, Nottingham University Hospitals NHS Trust, Queens Medical Centre Campus, Level E, East Block, Derby Road, Nottingham NG7 2UH, UK Received 4 February 2009; accepted 11 May 2009 Available online 30 June 2009
KEYWORDS Dysplastic kidney; PUV; VUR; Ultrasound
Abstract Objectives: Dysplastic kidneys (DK) are a common cause of chronic kidney disease (CKD). Little is known about their growth or how their sonographic appearance changes. This study aimed to test the hypothesis that DK gain little length, and to identify radiologic trends predictive of CKD. Methods: Ultrasound scans of children with DK born in 1980e2005 and referred to a single tertiary centre were analysed by a pediatric radiologist. Renal lengths were plotted on standard nomograms and the degree of dysplastic appearance noted. Factors related to DK e bladder outlet obstruction, vesico-ureteric reflux and renal impairment e were noted. Results: Fifty-three children were studied (83 kidneys), of whom 41 were boys; 289 scans were analysed. In 33 children there was associated bladder outlet obstruction or vesico-ureteric reflux. Forty-four DK were noted to fall off their renal length ‘centile’. This correlated well with the development of CKD and is statistically significant. Deterioration occurred in 53% of DK; primarily progressive reduction in corticomedullary differentiation. This also correlated well with development of CKD. Conclusion: More than half of the DK showed poor growth velocity. This, together with the degree of sonographic abnormality, carries a high predictive value for development of CKD. We recommend diligent serial sonography to follow renal growth and dysplastic appearance in children with DK. ª 2009 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved.
Introduction Dysplastic kidneys (DK) are an important cause of chronic kidney disease (CKD) in children [1]. They may arise de novo
* Corresponding author. Tel.: þ44 115 924 9924x62626. E-mail address: [email protected] (N. Fraser).
(primary) or be found in association with congenital obstruction of the urinary tract or vesico-ureteric reflux (VUR) (secondary). Certain genes, transcription and growth factors have been implicated in the development of DK but there is a paucity of information in the medical literature regarding the growth potential of these abnormal kidneys. This study set out firstly to test the hypothesis that DKs do not grow in length significantly, as measured by serial
1477-5131/$36 ª 2009 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jpurol.2009.05.014
Dysplastic kidneys in children postnatal ultrasonography. Secondly, we aimed to identify any sonographic features which were predictive for chronic renal failure.
41 interpretation of US (such as poor quality images, and some ectopic kidneys which were difficult to measure accurately).
Antenatal findings
Methods Case notes of all children with DK born between 1980 and 2005 referred to a single tertiary pediatric nephrourology centre were analysed. Sonographic criteria for the diagnosis of renal dysplasia were combinations of the presence of small-to-normal-sized kidneys showing echogenic parenchyma, reduction or loss of corticomedullary differentiation, and cortical cysts (Fig. 1). Postnatal ultrasound (US) studies were performed using a 5- and 7-MHz variable-focus transducer. All available renal tract US films were reviewed by a single senior radiologist in one institution. Longitudinal renal lengths were plotted against length or height of child on standard European nomograms (Fig. 2), the rationale being that published charts show an approximately linear relation between renal length and height in children over 1 year old [2]. This also avoids the additional variability of height with age, particularly important in children in our study population who may have CKD-related reduction in height velocity. In a very small number of cases film copies were unavailable, and reports of the US scans performed by the same radiologist or his colleagues were used. Children who had not undergone follow-up US were excluded, as were those whose scans were of poor image quality, as judged by the radiologist. Details drawn from the case notes included antenatal US findings, age at and mode of presentation, initial creatinine, and associated congenital abnormalities. The number of children with primary DK and those secondary to underlying pathology were established. The outcomes were plots of serial renal lengths recorded on nomograms compared to expected renal growth, and changes in dysplastic appearance. Also noted were survival, development of CKD and end-stage renal failure (ESRF). Assessment of degree of CKD was based on estimated GFR with reference to the Schwartz formula in most children, unless due to severe failure to thrive and poor body mass the creatinine did not reflect the child’s renal function. In these children, formal EDTA GFR was performed to attain a more accurate reflection of renal function. CKD was defined according to the National Kidney Foundation guidelines [3] (stages 1e5: stage 1, normal GFR but evidence of renal damage, stage 2 GFR 60e89, stage 3 GFR 30e59, stage 4 GFR 15e29, and stage 5 ESRF with GFR < 15 and on dialysis). For each renal US reviewed, the following parameters were recorded as well as renal length: age at first postnatal US, interval between subsequent scans, whether the DK were unilateral or bilateral, appearance of cortex (echogenicity, corticomedullary differentiation, presence of cysts, cortical loss or scarring), remainder of the urinary tract.
Results A total of 53 children were included in the study (83 renal units). Forty one (77%) of these were boys. Ten children were excluded due to incomplete data, i.e. difficulty in
In 32 children (60%), abnormal renal tracts were identified prenatally. Nine children had apparently normal renal tracts and a further three mothers were reported to have had a normal pregnancy. Results of fetometernal scan (FMS) were unrecorded in nine children. The majority of antenatal findings were first noted during the 20/40 fetal anomaly US, but some were picked up at 16/40, 18/40, 30/40 and 34/40. The commonest abnormalities were hydronephrosis (53%), small or discrepant renal size (34%) and oligohydramnios (34%). Other findings included echogenic cortex, cysts and ureteric dilation. PUV were suspected in seven boys. Thirty children with renal abnormalities on FMS underwent their first postnatal US at a mean of 35 days (range 1 daye8.5 months). The other two children presented later with symptoms. Both, in retrospect, had abnormalities on their antenatal scan. In the first child, antenatal bilateral hydronephrosis was seen but he was not referred until he developed a UTI at 5 weeks of age. In the second child, oligohydramnios was found during early FMS but this had settled during the pregnancy. He was not investigated until he presented 5 years later with enuresis.
Other modes of presentation Eleven out of the remaining 21 children presented during the neonatal period: one with UTI at 2 weeks of age, five with raised creatinine (two picked up on routine neonatal blood tests in premature babies), and three were screened due to other abnormalities (VACTERL, lower mesodermal defect sequence, two-vessel umbilical cord). Two boys were noted to have a palpable bladder and poor urinary stream during the first week of life. One of these was subsequently found to have PUV and the other neuropathic bladder. The remaining 10 children presented later with the following: UTI (6, at ages 5 weeks, 2 months, 5 months, 6 months, 15 months and 6 years), high creatinine (1), renal tract screening due to history of familial renal dysplasia (1, age 2 years), screening due to two-vessel umbilical cord at age 2 months (1), investigation for short stature at 4 years of age.
Primary and secondary DK The number of children with primary DK (defined as no evidence of underlying factors such as VUR, PUJ obstruction or bladder outlet obstruction) was nine. There were 33 children in whom an underlying cause could be identified, based on results of VCUG, postnatal US and other investigations. The commonest underlying factor was VUR, followed by PUV, PUJ obstruction and one each of bladder agenesis, neuropathic bladder and congenital urethral stricture. In the remaining 11 children with DK, it was not possible to clarify whether these were primary or secondary (Table 1).
Figure 1 Images 1 and 2: supine and prone scans of a normal right kidney. Renal cortex is less echogenic than adjacent liver and there is clear differentiation of cortex from pyramids in the medulla and central calyceal echoes. Image 3: small (3 cm) dysplastic right kidney on prone scan with echogenic cortex but some preservation of corticomedullary differentiation (CMD). Images 4 and 5: supine and prone scans of a good-sized dysplastic right kidney with echogenic cortex and loss of CMD, but no cysts. Image 6: small (2.5 cm) dysplastic right kidney on supine scan with echogenic cortex and loss of CMD, but no cysts. Image 7: small (3 cm) dysplastic right kidney on supine scan with echogenic cortex, loss of CMD, and several small cortical cysts. Image 8: two prone images of a good size (6.6 cm) dysplastic right kidney with echogenic cortex, loss of CMD, and multiple cysts e three in peripheral location posteriorly at the upper pole (right hand image), one in peripheral location anteriorly at the lower pole, and one deeper lower pole cortical cyst (left-hand image). Largest cyst, 6 mm maximum diameter.
Dysplastic kidneys in children
43 performed per child was three (range 2e6) and the mean interval between scans was 2 years (range 2 monthse6 years). Nineteen US scans were not available for analysis (reports used instead). Two were ectopic but measurements were possible. When the plots were analysed, 44 (53%) DK were noted to fall off their initial renal length ‘centile’, or fell further away from the -2 standard deviation (2 SD) line in the case of the very small DK. Ten of these kidneys could not be visualized on follow-up scans and were presumed to be very small. Nineteen grew along their initial plotted line. Twenty kidneys grew sufficiently to gain SDs.
Appearance of DK on initial US
Twenty-four children (44%) had at least one other congenital abnormality. These included syndromes and associations (5), other defects in the genitourinary tract, and miscellaneous anomalies. Twelve children with DK were born prematurely (20%), between 30/40 and 37/40 gestation [30/40 (1), 32/40 (3), 33/40 (2), 34/40 (1), 35/40 (1), 36/40 (3), 37/40 (1)]. IUGR was present in three children.
Four kidneys had so little cortex that the presence of increased echogenicity, reduced corticomedullary differentiation (CMD) and cysts was not possible to ascertain. The commonest sonographic feature indicative of renal dysplasia in the population studied was increased echogenicity (91%), followed by absent or reduced CMD (86%), and cortical cysts (30%). Worsening appearance was manifested by increasingly echogenic cortex, progressive reduction in CMD and appearance of new cysts. Deterioration (seen in 53%), and most commonly progressive reduction in CMD, correlated well with the development of renal insufficiency. This was statistically significant (95% CI of 35e71%). Conversely, of children maintaining normal renal function, most (92%) had either no change or improvement in renal appearance.
Laterality
Survival
Bilateral DK (32 children in all) were commoner than unilateral DK, the majority in boys. Right was slightly commoner than left; 44 and 39, respectively. In the nine children with primary DK, four were bilateral (44%). Of the children with secondary DK 70% were bilateral.
At the time of writing there were 52 survivors. One child with ESRF and Allagille syndrome died at the age of 13.5 months, but not as a direct complication of renal dysplasia.
Figure 2 Nomogram of renal length versus child’s height [2]. Red line demonstrates a kidney falling off its growth centile. Used with permission granted by Elsevier and Pediatric Radiology.
Congenital abnormalities and prematurity
CKD
Renal length in DK A total of 289 renal images were analysed and renal lengths measured and plotted on standard nomograms of renal length versus child’s height (Fig. 2). The mean number of US scans Table 1 Underlying (‘secondary’).
renal
tract
pathology
in
DK
Unilateral Bilateral Primary (n Z 9) 5 Secondary (n Z 33) VUR 8 PUV 0 PUJ obstruction 2 Neuropathic 0 bladder Urethral stricture 0 Bladder agenesis 0 Not ascertained 6 (n Z 11) Total: 53 21
4 11 8 1 1 1 1 5 32
The serum creatinine was normal throughout in a total of 13 children. Forty children (76%) had raised creatinine on presentation or shortly afterwards (above 120 mmol/L in 24) and in eight of these the creatinine normalized within 2 years. The remaining 31 surviving children went on to develop CKD. Three had EDTA GFR, the rest estimated GFR. Sixteen children developed mild renal impairment not requiring dialysis. The remaining 15 children have CKD or ESRF. Three children are currently receiving peritoneal dialysis or haemodialysis. Nine children have undergone renal transplantation, median age 6.4 years, seven of which were pre-emptive. Three children are awaiting transplant. One child’s transplant was complicated by venous thrombosis and he subsequently received a further deceased donor transplant. In children with CKD or ESRF four main patterns were noted: (1) bilateral DK: either both well below the 2 SD below mean line throughout, or one or both dropping in SD (commonest pattern, 34%) (Fig. 2);
44 (2) bilateral DK with good growth, no scarring but with underlying VUR or bladder outlet obstruction (28%); (3) solitary DK (19%): due to contralateral pathology e renal agenesis, multicystic dysplastic kidney, nephrectomy; half of these were growing, the other half had poor growth; (4) bilateral DK with good growth but global cortical loss or scarring (16%). Only one child with disappearing DK and a normal contralateral kidney developed chronic renal failure. All who have undergone or are awaiting renal transplantation had either bilateral DK with poor growth, solitary DK or scarred DK. Apart from renal impairment and its consequences, UTIs were a common finding (20 children) and were recurrent in 14 children.
Follow up Median duration of follow up was 7 years (range 1e18 years). All children with renal impairment have ongoing follow up with nephrologists. Ten with normal renal function were discharged and two lost to follow up.
N. Fraser et al. of the scans) review the US results. We accept that measuring both renal lengths sonographically and heights/ lengths in children can be difficult and may be a cause of inaccurate plotting on the nomograms. We recognise that over the time period US technology has improved; this is less likely to have impacted on renal measurements than renal appearance (the more subtle abnormalities may have been more readily detected in the latter years studied). There was a tendency to reject more of the earlier US results, where measurements were inaccurate or could not be estimated (usually US performed elsewhere, where low frequency probes were used, limiting spatial resolution), thus strengthening our data. We acknowledge that due to a relatively small number of US scans per patient our renal growth plots may merely indicate a trend. However, previous studies have shown that the observer variation in the measurement of renal length and size is equivalent to 2e3 years’ normal growth [4]. Therefore, more frequent scanning would not necessarily have been helpful over the time period studied. A longer duration, preferably prospective, study would be useful to confirm our findings.
Conclusion Discussion Dysplastic kidneys account for up to 14% of cases of CKD [1] and occur as a result of failure of normal fetal renal maturation. It is generally accepted that the diagnosis is reached sonographically, typified by small to normal sized, echogenic kidneys with reduction or loss of CMD, with or without cortical cysts. Histologically, dysplasia is characterized by the existence of primitive mesenchymal structures, cortical cysts and the primitive duct, the product of the ureteral bud system. Numerous studies have investigated the mechanisms involved in the development of renal dysplasia: the role of fetal urinary tract obstruction [5e7], the WNT-1 gene, PAX2, BCL2, galectin 3, transforming growth factor B1, and various types of transcription factors, cytokines and cell adhesion molecules [8,9]. To our knowledge, this is the first study which has objectively assessed renal growth in DK, correlating poor growth with the development of CKD. Furthermore, we have correlated worsening sonographic appearance with CKD. Overall, 53% of DK fell short of expected growth when plotted on a nomogram of renal length versus child’s height, despite some of them achieving an increase in length. This correlates well with the development of CKD and is statistically significant (44 out of 83 DK, 95% CI 42e64%). Solitary, poorly growing DK; bilateral, tiny DK/those dropping below the line of expected growth; scarred DK; and interestingly a group with bilaterally well growing DK with underlying PUV or VUR, were the main culprits for poor renal outcome in our series. More than half of the children going on to develop CKD independently had worsening dysplastic changes visible on serial US, which was also statistically significant. We eliminated inter-observer variation by having one experienced radiologist (who had himself performed most
As a result of our findings, we recommend that renal lengths of children with DK should be recorded on a standard nomogram, and a careful search attempted to pick up the subtle signs of worsening dysplastic appearance. Knowing that a fall-off in growth and worsening appearance both significantly raise the chances of developing CKD is important, and should prompt the clinician to ensure that close nephrological input is commenced and maintained.
References [1] Pistor K, Scharrer K, Olbing H, Tamminen-Mobius T. Children with chronic renal failure in the Federal Republic of Germany: primary renal diseases, age and intervals from early renal failure to renal death. Clin Nephrol 1985;23:278e84. [2] Dinkel E, Ertel M, Dittrich M, Peters H, Berres M, Sculte-Wissermann H. Kidney size in childhood. Sonographical growth charts for kidney length and volume. Pediatr Radiol 1985;15:38e43. [3] National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification and stratification. Am J Kidney Dis 2002;39(1):S1e266. [4] Sargent MA, Wilson BPM. Observer variability in the sonographic measurement of renal length in childhood. Clin Radiol 1992;46: 344e7. [5] Hoover DL, Duckett JW. Posterior urethral valves, unilateral reflux and renal dysplasia: a syndrome. J Urol 1982;128:994e7. [6] Shibata S, Shigeta M, Shu Y, Watanabe T, Nagata M. Initial pathological events in renal dysplasia with urinary tract obstruction in utero. Virchows Arch 2001;439:560e70. [7] Spencer JR, Maizels M. Inhibition of protein glycosylation causes renal dysplasia in the chick embryo. J Urol 1987;138:984e90. [8] Glassberg KI. Normal and abnormal development of the kidney: a clinician’s interpretation of current knowledge. J Urol 2002; 167:2339e51. [9] Yang SP, Woolf AS, Yuan HT, Scott RJ, Risdon RA, O’Hare MJ, et al. Potential biological role of transforming growth factor-B1 in human congenital kidney malformations. Am J Pathol 2000; 157(5):1633e47.