Pediatr Surg Int (1995) 11:18-21

© Springer-Verlag 1995

J. T h o r u p • K. L e n z • A. R a b o l • A. Passalides O. It. N i e l s e n

Follow-up of prenatally diagnosed unilateral hydronephrosis

Based on previous experience with prenatally diagnosed unilateral hydronephrosis, we found that the primary indications for surgical intervention should be symptoms or functional impairment of the hydronephrotic kidney. Nonoperative management of neonates without symptoms and with normal function of the affected kidney was proposed. However, the strategy of treatment after prenatally diagnosed hydronephrosis is still controversial. We studied 28 consecutive children with suspected unilateral pelviureteral junction obstruction and a normal contralateral kidney. The overall follow-up period varied between 2.5 months and 6 years (median 2 years). Eleven children had normal function of the hydronephrotic kidney and were managed nonoperatively throughout the follow-up period. None of these demonstrated any symptoms and the renal function remained normal. A further 4 children with normal function of the affected kidney were managed nonoperatively, but later had a pyeloplasty performed because of either symptoms or deterioration of renal function. Eleven children had a pyeloplasty performed after the first renography showed that the hydronephrotic kidney provided less than 40% of total renal function. The age at pyeloplasty was 3 weeks-7 months (median 6 weeks). In all cases but 1 the function of the affected kidney improved. Two patients with impaired hydronephrotic renal function were not operated upon. Our results indicate no need to change the strategy of treatment. Abstract

Kidney • Hydronephrosis • Ureteral obstruction • Renal function tests K e y words

J. Thorup (~3) • K. Lenz • A. Rabol • A. Passalides • O. H. Nielsen The Department of Paediatric Surgery 4062 and The Department of Nuclear Medicine, State University Hospital, DK-2100 Copenhagen, Denmark

Introduction One decade has passed since we evaluated our experience with the first nine cases of prenatally diagnosed unilateral hydronephrosis [13]. Based on the results of that study, we proposed that the primary indications for surgical intervention should be symptoms or functional impairment of the hydronephrotic kidney. Due to the risk of operative complications, we found it unacceptable to operate upon neonates without symptoms and with normal function of the affected kidney [13]. Other authors found it difficult to accept that anyone could observe a kidney known to be obstructed and wait for symptoms to develop before intervening [1]. Unilateral renal obstruction should be corrected as soon as it can be diagnosed in neonates and very young infants, as otherwise contralateral hypertrophy will occur and reduce the potential for recovery of function in the obstructed kidney [7]. According to the literature, the results of early postnatal pyeloplasty are excellent [2, 4, 6, 7, 9]. However, the results of a nonoperative approach to 45 neonates with unilateral hydronephrosis have recently been published [8]. The natural history of untreated unilateral hydronephrosis as described in this series provides reassuring evidence that a nonoperative approach, with close and careful follow-up, does not appear to place the kidney or the patient at an unnecessary risk and can be safely continued until obstruction is defined or excluded. The widespread use of maternal ultrasonography (US), which permits fetal hydronephrosis to be detected easily, has created a virtual epidemic of asymptomatic infants with this condition. Since the strategy of management still seems controversial, we find it important to give information on the cases treated since we published our initial experience [13l.

19 Affected Kidney GFR (% of total) 605040 30 2010 0

t'o

0

2'0

'

~0

45

'

go

65

Age (months)

Fig. 1 Percentage of renal function contributed by hydronephrotic kidney compared to patient age. Solid lines connect sequential measurements in same patient. All patients had nonoperative management throughout follow-up period (GFR glomemlar filtration rate)

structed and the individual renal function was calculated from the tracer uptake in the parenchyma from 80 to 130 s after background subtraction. Sequential images were collected at 2-rain intervals to evaluate gross anatomy. Prior to the examination, the children were hydrated with i. v. Darrow-glucose solution 5 ml/kg for 4 h. Evaluation of drainage was not included in the study. Estimation of overall renal function (TRF) was made by serum creatinine and blood urea concentrations as well as chromated Cr 51ethylenediaminetetra-acetic acid clearance. Scintigraphy, renography, and US were repeated regularly during follow-up. In the patients operated upon, a Hynes-Anderson pyeloplasty was performed. Atraumatic monofiber absorbable 5-0 inte~cupted sutures were used for anastomosis. After the pyeloplasy, a 5 Fr feeding tube was left indwelling for a few days to serve as a nephrostomy. No internal stent or drain was used. Another 7 patients with suspected unilateral PUJ obstruction were seen within the study period after prenatal diagnosis, but excluded. One had a percutaneous nephrostomy performed before the initial renographic evaluation was carried out; 6 had postnatal US showing unilateral pelvic dilatation of 10-36 mm and renography showing normal function of the hydronephrotic kidney. Five of these patients had no IVP to exclude a megaureter and confirm the normal state of the contralateral kidney, and 1 had no follow-up investigations.

Materials and methods The study included 28 consecutive children (10 girls, 18 boys) with suspected unilateral pelviureteral junction (PUJ) obstruction (8 rightsided, 20 left-sided) and a normal contralaterai kidney. All but 2 had had prenatal US showing hydronephrosis. In 2 cases hydronephrosis was diagnosed by US on the 1st day of life at routine examination for congenital malformations. A correct diagnosis was obtained by performing repeated postnatal US, i. v. pyelography (IVP), and renal scintigraphy. All patients had a normal urinary bladder; none had vesicoureteric reflux or associated malformations. Renal function was evaluated with technetium Tc99m MAG-3 scintigraphy and renography. Scintigraphy was performed with a gamma-camera (General electric Gammacamera 400 AC) using a posterior view that included the kidneys, ureters, and bladder, and 120 frames 10 s in duration were stored in a computer (General Electric Starcam 3000 computer). Time-activity curves were c o n -

Fig. 2 Percentage of renal function contributed by hydronephrotic kidney compared to patient age. Solid lines connect sequential measurements in same patient. All patients had early pyeloplasty. First measurement for each patient indicates preoperative hydronephrotic kidney function (GFR glomemlar filtration rate) Affected Kidney GFR

(% of total)

6°t 504030. 20100 0

t'o

2~

3~

4'o Age (months)

Results All 28 cases included in the study had an IVP showing significant distension of the calices and dilatation of the renal pelvis on the affected side. The overall follow-up period varied between 2.5 months and 6 years (median 2 years). In 13 cases indications for immediate operative treatment were found after postnatal renographic evaluation (i. e., the hydronephrotic kidney provided < 4 0 % of TRF). However, in 2 cases the parents did not accept an operation although the function of the hydronephrotic kidney was shown to be impaired. Eleven children had normal function of the hydronephrotic kidney and were managed nonoperatively throughout the follow-up period. The results of renographic evaluation of the function of the hydronephrotic kidney in percentage of T R F for the 13 patients not operated upon are shown in Fig. 1. None of the patients received prophylactic antibiotics, and none of the children demonstrated any symptoms during the follow-up period. Eleven children had a pyeloplasty performed after the first renography. The results of pre- and postoperative renographic evaluation of hydronephrotic kidney function in percentage of T R F are seen in Fig. 2. None of the children had any symptoms before the operation. The median age at the time of pyeloplasty was 6 weeks (range 3 w e e k s - 7 months). In 7 cases the obstruction was produced by an S-shaped insertion of the ureter with mechanical compression from adhesive bands, in 3 by poor transmission of peristalsis through a narrow segment, and in 1 there was a stricture. There were no postoperative complications. Postoperative ward time varied between 2 and 11 days (median 7 days). During the follow-up period 1 patient had one episode of urinary tract infection; the others presented no symptoms.

20 Affected Kidney GFR (% of total)

Fig. 2, the change in renal function from the preoperative renographic evaluation to the last renography at follow-up showed a mean percentage improvement of 37% (range 0%-175%). These figures are in agreement with others, who have reported significant improvement of hydronephrotic kidney function ( > 10% of TRF) in all patients after so ~ very early pyeloplasty [4, 6]. Recently, Koff and Campbell [8] followed 15 neonates with severe hydronephrosis nonoperatively, all having an 30 obstructed diuretic renogram and markedly decreased hydronephrotic kidney function. During follow-up relative 20 and absolute renal function rapidly increased in all patients, 10 hydronephrosis improved in 7, and contralateral compensatory hypertrophy did not develop in any case. However, the duration of follow-up was short ( 2 - 2 2 months). Within this o ~'o 4'0 ~o ' Age (months) period of follow-up, 2 of our patients followed nonoperatively with impaired function of the hydronephrotic kidney Fig. 3 Percentage of renal function contributed by hydronephrotic also showed improved relative renal function (10% each, kidney compared to patient age. Solid lines connect sequential measurements in same patient. Stars indicate hydronephrotickidney func- Fig. 1). Although they are only 2 cases, the increase in hydronephrotic kidney function was less impressive comtion at time of pyeloplasty (GFR glomerular filtration rate) pared to the cases operated upon early. The same laboratory facilities were available for all our patients, making the Figure 3 shows the renographic results of 4 children with results of the nonoperated cases comparable to those of the normal function of the hydronephrotic kidney, initially operated ones. Ransley et al. [12] studied 27 hydronephrotic kidneys managed nonoperatively but later requiring a pyeloplasty because of either symptoms or deterioration of renal func- with moderate function, including those with differential tion. In 2 cases obstruction was produced by an S-shaped uptake function on a 99mTc-DTPA scan of 20%-39%. None insertion of the ureter with mechanical compression from of these had a pyeloplasty performed before 3 months of adhesive bands and in 2 by poor transmission of peristalsis age. Of 23 operated kidneys, 9 showed no change in uptake through a narrow segment. After operation, the function of function postoperatively despite a decrease in hydronephrothe hydronephrotic kidney increased slightly in 2 cases and sis and improved drainage. These results are not much better than those seen after pyeloplasty in older children decreased slightly in 2. [13, 14]. Thus, the potential for at least partial recovery of hydronephrotic kidney function may be lost for some children with moderate functional impairment as early as 3 months postnatally. It is generally accepted that if the Discussion obstructed kidney provides less than 10% of TRF postnaCompared to other studies, we observed a high proportion tally, it will only rarely respond positively to repair of the of children with congenital impairment of the function of obstruction or decompression [4, 10, 12]. To date, 4 of 15 postnatally normally functioning kidthe hydronephrotic kidney. In 46% of the cases the hydronephrotic kidney function was < 4 0 % of TRF at the first neys ( > 40% of TRF) in our study have required pyeloinvestigation. The figures from the literature are 2 9 % - 3 3 % plasty. These figures are in accordance with results from the [8, 10, 12], however, selection is important. Due to the lack literature [10, 12]. Two of our deteriorating kidneys showed of IVP and follow-up by renography, we excluded 6 transient further impairment after operation (Fig. 3). The deteriorating kidney may be especially vulnerable to surpossibly less severe cases from our study. In the present investigation, we dealt only with cases of gery: it has been shown that postoperative gain after significant hydronephrosis caused by suspected impaired function in adult hydronephrotic kidneys may be PUJ obstruction after US and radiologic assessment. We a process of long duration [3]. Transient edema at the level made no attempt to interpret the initial radionuclide wash- of the anastomosis might, of course, explain the renoout kinetics with regard to obstruction. According to our graphic findings. However, similar findings were not seen present knowledge, all ureters seemed obstructed, but we in the cases with immediate postnatal surgery, which makes agree with other investigators that the tests generally used it more logical to interpret them as coincidental. All our in older children do not appear valid where the accurate operated patients were scheduled to have the first postdiagnosis of obstruction in the neonate is concerned [4, 8, operative renography performed after 1 month; at that time 12]. When renographic findings are considered in the functional loss due to iatrogenic obstruction is probably still present study, they refer specifically to uptake function reversible [5]. A nonoperative approach requires very close and careful and, therefore, individual glomerular filtration rate. In all but 1 case the function of the hydronephrotic follow-up during childhood. In this respect, it is noteworthy kidney improved after immediate pyeloplasty. According to that we did not find aberrant vessels to be the cause of

oo1

21 obstruction in any of our operated cases. Aberrant vessels as the cause of PUJ obstruction are seen more often in adults [11]. The morphologic changes related to hydronephrosis caused by aberrant vessels m a y be modest before symptoms occur. A large number o f children with normal function of the hydronephrotic kidney and nonoperative m a n a g e m e n t will need many follow-up investigations. However, in some of these cases the hydronephrotic kidney will probably lose function anyway. Provided that the operative procedure is safe, i m m e d i a t e postnatal surgery on hydronephrotic kidneys with moderately impaired function ( 1 1 % - 4 0 % of TRF) will lead to i m p r o v e m e n t of function in all cases. With nonoperative m a n a g e m e n t the potential for recovery will probably be less. Our results do not indicate any change of our previously proposed strategy of treatment [13]. However, controlled randomized studies are required in order to improve interpretation of renographic results and determine whether certain subgroups will benefit specifically from early pyeloplasty.

References 1. Allen TD (1985) Editorial comments. J Urol 134:917 2. Bernstein GT, Mandell J, Lebowitz RL, Bauer SB, Colodny AH, Retik AB (1988) Ureteropelvic junction obstruction in the neonate. J Urol 140:1216-1221

3. Djurhuus JC, S~rensen SS, JCrgensen TM, Taageh~j-Jensen F (1985) Predictive value of pressure flow studies for the functional outcome of reconstructive surgery for hydronephrosis. Br J Urol 57:6-9 4. Hanna MK, Gluek R (1988) Ureteropelvic junction obstruction during the first year of life. Urology 31:41-45 5. Holm-Nielsen A, JCrgensen T, Mogensen P, Fogh J (1981) The prognostic value of probe renography in ureteric stone obstruction. Br J Urol 53:504-507 6. King LR, Coughlin PWF, Bloch EC, Bowie JD, Ansong K,. Hanna MK (1984) The case for immediate pyeloplasty in the neonate with ureteropelvie junction obstruction. J Urol 132:725-728 7. King LR, Hatcher PA (1990) Natural history of fetal and neonatal hydronephrosis. Urology 35:433-438 8. Koff SA, Campbell K (1992) Nonoperative management of unilateral neonatal hydronephrosis. J Urol 148:525-531 9. Koyle MA, Ehrlich RM (1988) Management of ureteropelvic junction obstruction in neonate. Urology 31:496-498 10. Madden NP, Thomas DFM, Gordon AC, Arthur RJ, Irving HC, Smith SEW (1991) Antenatally detected pelviureteric junction obstruction. Is non-operation safe? Br J Urol 68:305-310 11. Poulsen EU, Frokjaer J, Taagehj-Jensen F, JCrgensen TM, Norgaard JE Hedegaard M, Djurhuus JC (1987) Diuresis renography and simultaneous renal pelvic pressure in hydronephrosis. J Urol 138:272-275 12. Ransley PG, Dhillon HK, Gordon I, Duffy PG, Dillon MJ, Barratt TM (1990) The postnatal management of hydronephrosis diagnosed by prenatal ultrasound. J Urol 144:584-587 13. Thorup J, Mortensen T, Diemer H, Johnsen A, Nielsen OH (1985) The prognosis of surgically treated congenital hydronephrosis after diagnosis in utero. J Urol 134:914-917 14. Valayer J, Adda G (1982) Hydronephrosis due to pelviureteric junction obstruction in infancy. Br J Urol 54:451--456