- Email: [email protected]
Infant pyeloplasty is a low-risk procedure
Infant Pyeioplasty By Donald
Is a Low-Risk
B. Shaul, James A. Cunningham, Phillip Lowe, Leonard Los Angeles, California
0 Early pyeloplasty for the treatment of congenital ureteropelvic junction obstruction to maximize nephron salvage is only justified if the potential hazards of operating on small infants are avoided. The records of all infants who underwent pyeloplasty by the authors over a 7-year period were analyzed. The outcome of surgery in the younger infant group (YIG; patients less than 2 months of age) was compared with that of the older infant group (OIG; patients more than 2 months of age). Preoperative evaluation in cases of mild or moderate hydronephrosis was directed toward ruling out a nonobstructed collecting system and included voiding cystourethrography and serial ultrasonography and/or dual isotope diuretic renography. Open pyeloplasty was performed if collecting systems had deteriorated or were demonstrated to be obstructed; it was also performed for severe cases of hydronephrosis. Postoperative assessment consisted of serial ultrasonography and/or nuclear imaging to confirm decompression and relief of obstruction. Thirty three pyeloplasties were performed in 31 patients in the YIG (two bilateral), and 33 were performed in 32 infants in the OIG (one bilateral). The only significant differences between the groups were as follows. Patients in the YIG were more likely to present in utero (87% Y 53%; P < .Ol), whereas those in the OIG were more likely to present with a urinary tract infection (44% Y 0%; P < .Ol). The YIG was more likely to have nephrostomy drainage postoperatively (67% 1145%; P < .Ol). In the overall series, there were five complications, all of which were postoperative infections requiring intravenous antibiotic therapy. Two occurred in the YIG and three in the OIG. One patient had a urinoma, which was drained percutaneously. During follow-up examination, 32 renal units in the YIG and 30 in the OIG were judged to be stable or improved. One kidney in each group was not salvaged after the pyeloplasty. Two patients in the OIG were lost to follow-up. No anesthetic complications or strictures occurred in either group. Good results of pyeloplasties performed in the infants in this series support early correction of ureteropelvic junction obstruction in infants. Copyright c 1994 by W.B. Saunders Company INDEX WORDS: Hydronephrosis, junction obstruction.
pyeloplasty;
ureteropelvic
I
NCREASES in the routine usage of prenatal ultrasonography has led to the diagnosis of hydronephrosis in utero in a large number of pregnancies. Ureteropelvic junction (UPJ) stenosis was found by prenatal ultrasound in 0.1% of pregnancies.’ The postnatal management of these dilated, potentially obstructed neonatal collecting systems is controversiai. McCrory et al were the first to demonstrate that relief of obstruction prevents further renal damage and, in a subgroup of infants, allowed for some functional recovery.’ Results of animal studies have also suggested that early relief of obstruction is JournalofPediafric Surgery, Vol 29, No 2 (February), 1994: pp 343-347
Procedure B. Skaist, and Brian E. Hardy
beneficial, even in cases of partial ureteral obstruction.3,4 This has led some investigators to encourage early pyeloplasty in cases of UPJ obstruction initially diagnosed antenatally. However, other investigators have been more selective and maintain that even severe hydronephrosis is not necessarily obstructive in nature.5-7 They have promoted observation with serial radiological evaluation for the management of unilateral neonatal hydronephrosis. The final functional outcome of these hydronephrotic, potentially obstructed kidneys remains to be seen. There is little doubt that early pyeloplasty in definite cases of UPJ obstruction will yield the best long-term functional result. In cases of significant hydronephrosis secondary to UPJ stenosis, it is also likely that pyeloplasty, when carried out with minimal complications, will provide the best ultimate outcome. This study was undertaken to determine whether there is a significant difference in postoperative results between young infants and older infants. MATERIALS
AND METHODS
The records of all infants undergoing pyeloplasty by the authors (between 1985 and 1992) were reviewed. Both inpatient and outpatient records were reviewed, and data were collected regarding patient demographics, associated anomalies, preoperative evaluation, hospital course, postoperative evaluation, and outcome. Patients were divided into two groups. based on age at the time of surgery. Patients undergoing pyeloplasty before 2 months of age formed the younger infant group (YIG), whereas those undergoing pyeloplasty at more than 2 months of age were in the older infant group (OIG). Preoperative evaluation was directed toward ruling out nonobstructive causes of hydronephrosis. Voiding cystourethrography (VCUG) was performed to rule out vesicoureteral reflux. All patients underwent confirmatory postnatal ultrasound. If severe or massive hydronephrosis was present with thinning of the renal parenchyma, pyeloplasty was performed. However. if mild or moderate hydronephrosis was present, serial ultrasonography was performed 3 to 6 weeks later. If deterioration was noted, the patient underwent pyeloplasty. If there was no change, dual isotope diuretic renography was performed to determine whether obstruction was present. Agents used included technetium 99m-diethylenetriaminepentaacetic acid (Y’hnT~-DTPA) and iodine 131 orthoiodo Hippuran (i3iI-OIH). Furosemide (0.5 mgikg) was administered 16 minutes after injection of the tracer.
From the Division of Pediatric Urology, Children’s Hospital, Los Angeles, CA. Presented at the 24th Annual Meeting of the American Pediatric Surgical Association, Hilton Head, South Carolina, May 15-18, 1993. Address reprint requests to Brian E. Hardy, MD. 1300 North Vermont Ave. Suite No. 708, Los Angeles. CA 90027. Copyright 0 I994 by KB. Saunders Cornpan? 0022-3468/2902-0040$03.0010 343
SHAUL ET AL
RESULTS
Fig 1. The stent is created by cutting extra side holes in a pediatric feeding tube. It is important to position the catheter so that all the side holes are in the pelvis or proximal ureter.
Patients treated in the last 5 years of the study period had diuretic-induced isotope washout curves constructed. In patients with good function (renal isotope excretion), the time to wash out half the isotope was measured and recorded as the tl/2. t’/r values of greater than 30 indicated obstruction. Cases of obstruction demonstrated by this method also underwent pyeloplasty. Patients whose kidneys showed improvement on serial ultrasonography, partial obstruction on nuclear medicine studies, or no obstruction did not undergo surgery and are not included in this series. In all patients the kidney was exposed via an extraperitoneal flank approach. The type of reconstruction used was the AndersonHynes dismembered pyeloplasty. Resection of redundant pelvis was performed only in cases of massive dilatation. Drainage of the bladder was routine for 2 to 4 days postoperatively. Patients were more likely to have nephrostomy drainage and/or stent placement when there was perirenal inflammation, recent infection, a solitary functioning kidney, or massive dilatation of the pelvis with calyces that were likely to be filled with debris. In the majority of patients a nephrostomy tube stent combination was fashioned using a 5F feeding catheter with multiple extra side holes (Fig 1). The catheter was placed such that the tip lay beyond the anastomosis in the proximal ureter, and the extra side holes lay within the pelvis of the kidney. The tube was brought out through the inferiormost calix, and a purse-string suture was placed in the renal capsule surrounding the tube to prevent extravasation around the tube. The tube was left plugged, but was unplugged if a nephrostomy was needed. Penrose drainage of the perirenal space was routinely employed, with the penrose brought out through a separate stab incision. Antibiotic prophylaxis was continued in the postoperative period for an average of 2 weeks. Postoperative evaluation consisted of serial ultrasonography. Progressive decrease in the amount of dilatation of the renal pelvis was equated with successful relief of obstruction. If sequential ultrasonography showed hydronephrosis or if there were concerns about the functional status of the kidney, the patients also underwent dual isotope diuretic renography to confirm that obstruction had been relieved. When possible, follow-up examinations were performed yearly, for up to 4 years postoperatively, to detect recurrence of UPJ obstruction or development of contralateral UPJ obstruction. Statistical analysis was performed using a statistical program (Stat View, Calabasas, CA). Normally distributed continuous variables were analyzed using the Student’s t test for unpaired data. Other continuous variables were analyzed using the MannWhitney U test. The ,$ test was used to analyze noncontinuous variables.
Thirty three pyeloplasties were performed in 31 infants in the YIG, and 33 were performed in 32 infants in the OIG. Three infants underwent bilateral pyeloplasty, two of which were performed under the same anesthetic (Table 1). There was male predominance in both groups. The left side was involved approximately twice as often as the right side. The mean age of YIG patients was 36 + 16 days, and the mean age of OIG patients was 160 + 91 days. There was a significant difference in the mode of presentation between the groups (Table 2). Patients in the YIG were more likely to present as the result of a prenatal ultrasound finding than were those in the OIG (87% v 53%; P < .Ol). Patients in the OIG were more likely to present with a urinary tract infection (44% v 0%; P < .Ol). In the total series, five patients presented with an abdominal mass. One patient with a contralateral multicystic dysplastic kidney presented with acute renal failure secondary to the UPJ obstruction. Associated anomalies were most likely to have involved the genitourinary system (Table 3). Five infants had a contralateral multicystic dysplastic kidney (11%). All these patients were in the YIG when early decompression of their contralateral obstruction was performed. Vesicoureteral reflux occurred in four patients (7%). In two instances, the reflux was into the obstructed unit, and in two it was into the contralateral kidney. Four infants had contralateral hydronephrosis. These cases were evaluated with dual isotope diuretic renography. Two were partially obstructed, and two were not obstructed. In all four the hydronephrosis resolved without operative treatment. Studies used in the preoperative evaluation of the patients are shown in Table 4. Voiding cystourethrography (VCUG) was performed in 89% of the patients preoperatively. Patients who did not have preoperative VCUG were those for whom the diagnosis of obstruction was clearly established by other methods of investigation. With the exception of one patient for whom the diagnosis was made preoperatively from an incidental finding on a computed tomography scan, all patients underwent ultrasonography. Diuretic renography was performed preoperatively to quantitate Table 1. Patient Demographics Young
Age No. of patients No. of pyeloplasties No. of bilateral pyeloplasties
Infants
Older Infants
<2mo 31
>2mo
33
33
2
32 1
Sex (male:femalel
22:9
23:9
Side (right:left)
12:17
8:23
345
INFANT PYELOPLASTY
Table 4. Preooerative Evaluation
Table 2. Mode of Presentation Young
Older
Infants
Young Infants
Infants
In = 31)
(n = 3.2)
(n = 311
Older
Infants
In = 321
27
17*
Voiding cystourethrography
28
28
Urinan/ tract infection
0
14’
Renal ultrasound
30
32
Mass
2
3
Acute renal failure
1
0
Prenatal ultrasound
t% washout
NOTE. Patients may have more than one mode of presentation. *P i
21
23
70.4 + 33
64 r 28*
15
18
0
1
isotope renal scan
.Ol.
Intravenous
(min) pyelography
Whitaker test *P is not significant.
the degree of obstruction in 44 patients (70%). Values for t’/z were 70.4 -+ 33 minutes in the YIG and 64 +- 28 minutes in the OIG. These values were not significantly different. Intravenous pyelography was performed on 33 patients, and one patient had a Whitaker test preoperatively. Five patients had lower urinary tract infections postoperatively (two in the YIG, three in the OIG). All resolved with antibiotic therapy. One patient had a urinoma, which was drained percutaneously and did not recur. Other postoperative results are shown in Table 5. Younger infants were more likely to have their anastomosis protected by nephrostomy drainage and/or stent placement than were older infants. The length of postoperative stay was not different between the groups. Patients in the YIG had significantly longer follow-up than did those in the OIG (23 & 21 months v 12 + 12 months). Seventy-three percent of the patients in each group had follow-up for at least 3 months. Two patients in the OIG did not return for postoperative studies; the 64 remaining renal units had between one and five postoperative ultrasounds, with an average of 2.5 postoperative ultrasounds per patient in each group. Improvement of hydronephrosis was noted for 30 renal units in the YIG and 29 units in the OIG. Three patients did not return for additional follow-up studies after an early postoperative ultrasonography demonstrated no change in the hydronephrosis. Although many cases with this finding had improvement subsequently, the ultimate outcome of these three patients is not known. Nine patients had postoperative isotope scans; there was Table 3. Associated Genitourinary
Vesicoureteral
Older
Infants
Table 5. Postoperative
In = 31)
h = 32)
5
0
1
1
Postoperative
Contralateral
0
2
Follow-up
0
2
1
1
Crossed ectopia
1
0
Improved
drainage
stay (d)
(mo)
No follow-up
nonobstructive
hydronephrosis
(n = 33)
Nephrostomy/stent
partial UPJ
obstruction
Results
Young Infants
reflux
lpsilateral
Contralateral
Changes in renal function after relief of urinary tract obstruction in pediatric patients was initially studied by McCrory et al.’ Creatinine clearance, urinary concentrating ability, and excretion of an acid load were studied in 15 patients before and after treatment of urinary tract obstruction. These investigators noted that infants have the greatest capacity for improvement after relief of obstruction. It is well known that renal function continues to mature after birth.6 Elimination of potentially harmful obstruction from these maturing nephrons makes sense intuitively. The benefit of early decompression has been applied to large groups of neonates and infants, with good results.8-1”Simultaneously, experimental analyses were performed in different animal models.3,4 A series of experiments in a neonatal guinea pig model further elucidated the effects of permanent and temporary partial unilateral ureteral obstruction on renal function. It was noted that even though the intraureteral pressures were returned to normal after relief of obstruction, renal functional impairment persisted into adulthood.14 Even though these carefully performed studies appear to show the harmful effects of untreated unilateral ureteral obstruction on renal function, the treatment of neonatal hydronephrosis in humans has remained controversial.
dysplastic
Kidney (contralateral)
Contralateral
DISCUSSION
Anomalies
Young Infants
Multicystic
normal function in three, adequate function (but a small kidney) in four, and no function on the involved side in two. The latter six patients had markedly dilated collecting systems preoperatively.
studies
Older
Infants
(n = 33)
22
15*
7.7 f 6.1
5.8 f 5.0
23 f 21
12 f 12*
0
2
30
29
Stable
2
1
Not salvaged
1
1
lP < .Ol.
346
SHAUL ET AL
The appropriate treatment for neonates diagnosed in utero as having hydronephrosis secondary to anomalous UPJ was addressed by Homsy et aL5 Twenty-one of 60 renal units were believed to be partially obstructed or nonobstructed, despite being hydronephrotic. Because deterioration occurred in only 41% of patients not treated operatively, the investigators recommended a period of observation for all patients in whom obstruction could not be clearly demonstrated. Similar nonoperative approaches have been adopted by other investigators.6,7 A consistent conclusion among these investigators is that a definite diagnosis of obstruction is difficult to make. For example, the Whitaker test was found to be unreliable in the diagnosis of neonatal ureteral obstruction.’ In another series, all the standard modalities used to detect obstruction were unreliable when applied to neonates. It was suggested that contralatera1 compensatory hypertrophy is the most sensitive indicator of obstruction.6 The combination of evidence demonstrating the harmful effects of untreated obstruction on neonatal kidneys and the difficulty one faces in proving obstruction makes the approach of observing potentially obstructed neonatal kidneys worrisome. If UPJ obstruction can be safely relieved in these young infants, the potential damage that occurs during the observation period may be avoided. The safety with which pyeloplasty can be performed in neonates and young infants has been demonstrated.8-1”$‘5 Sixty-six pyeloplasties were performed in 63 infants in our series. Roughly half these procedures were performed in infants less than 2 months of age. Despite the young age and small size of the infants, there was no appreciable difference in outcome between this group and the OIG. There were no anesthetic complications and no deaths in the series. Follow-up was inadequate in five cases. Of the remain-
ing kidneys, postoperative ultrasonography demonstrated improvement in all cases except some that had a severe degree of dilatation preoperatively. Performance of pyeloplasty in the YIG did not prolong the hospital stay. Younger patients were more likely to have their repairs protected by some combination of nephrostomy and/or stent placement. These tubes were always removed 1 to 2 weeks after surgery. This type of drainage is considered to be a safeguard rather than a necessity. The fact that all tubes were able to be removed within 2 weeks confirms that significant anastomotic obstruction did not occur. There is no evidence that placement of these tubes promotes anastomotic obstruction, as has been suggested.6 The longer follow-up in the YIG is a reflection of parental compliance. Early postnatal treatment was sought for more than 80% of infants in the YIG. Even though nearly 50% of OIG cases were diagnosed in utero, most parents did not seek treatment for their infants until several months after delivery. This pattern continued postoperatively. It is interesting to note that 44% of the patients in the OIG presented with a lower urinary tract infection. The same finding has been noted by others.711” The combination of urinary stasis in the dilated upper tract as well as infection in the lower tract creates the possibility of a serious urinary infection. For this reason, urinary prophylaxis has been recommended until dilatation is relieved.‘” Unfortunately, the definitive diagnosis of UPJ obstruction is often difficult to make. The relief of obstruction by resection of the UPJ and performance of a dismembered pyeloplasty has now been safely achieved for large groups of infants. We believe that prolonged observation of those kidneys, which are likely to be obstructed, is currently unjustified.
REFERENCES 1. Grignon A, Filiatrault junction stenosis: Antenatal investigation, and follow-up.
D. Homsy Y, et al: Ureteropelvic ultrasonographic diagnosis, postnatal Radiology 160:649-651. 1986
2. McCrory WW, Shibuya M, Leumann E, et al: Studies of renal function in children with chronic hydronephrosis. Pediatr Clin North Am l&445-465, 1971 3. Chevalier RL, Kaiser DL: Chronic partial ureteral obstruction in the neonatal guinea pig. Influence of uninephrectomy on growth and hemodynamics. Pediatr Res l&1266-1271, 1984 4. Taki M, Goldsmith DI, Spitzer A, et al: Impact of age on effects of ureteral obstruction on renal function. Kidney Int 24:602-609, 1983 5. Homsy YL, Williot P, Danais S: Transitional neonatal hydronephrosis: Fact or fantasy? J Ural 136:339-341, 1986 6. Koff SA, Campbell K: Nonoperative management of unilateral neonatal hydronephrosis. J Urol148:525-531,1992 7. Dhillon
HK,
Duffy
PG,
Gordon
I. et al: A ramdomised
clinical trial in infants with prenatally diagnosed unilateral hydronephrosis. Presented at the meeting of the Section on Urology of the American Academy of Pediatrics. San Francisco. CA. October 12, 1992 8. Bevani B, Belman BA: Ureteropelvic newborns and infants. J Urol 128:770-773.
junction 1982
obstruction
9. Roth PR. Gonzales ET: Management of ureteropelvic tion obstruction in infants. J Ural 129:108-110, 1983
in
junc-
10. King LR, Coughlin PWF, Bloch EC, et al: The case for immediate pyeloplasty in the neonate with ureteropelvic junction obstruction. J Urol 132:725-728. 1984 11. Murphy JP, Holder TM, Ashcraft KW, et al: Ureteropelvic junction obstruction in the newborn. J Pediatr Surg 19:642-648, 1984 12. Flake AW, Harrison MR, Save L, et al: Ureteropelvic junction obstruction in the fetus. J Pediatr Surg 21:1058-1063, 1986
INFANT PYELOPLASTY
347
13. Wolpert JJ, Woodard JR, Parrott TS: Pyeloplasty in the young infant. J Urol 142573-575. 1989 14. Chevalier RL: Renal response to ureteral obstruction in early development. Nephron 56:113-l 17, 1990
15. Bernstein GT, Mandell J. Lebowitz RK, et al: Ureteropelvic junction obstruction in the neonate. J Ural 140:1216-1221.1988 16. Koff SA: Presented at the Los Angeles Urological Society, September 1992
Discussion In reviewing your manuscript, Dr Sham, several questions come to mind. First, it appears that you selected your patients on the basis of severe dilation, and the ones that were the most hydronephrotic were the ones that were operated on. Also, patients whose renal scan showed decreased function or delayed drainage were operated on too. It was a little unclear as to the number of patients that were observed nonoperatively, and the method of follow up. If you could clarify that for us, it would help. Second, do the conclusions of your report indicate that you are going to treat these kidneys differently in the future? Are you going to be more aggressive with operating on them, or are you going to manage them in the same way? If you could give us an idea of your algorithm on how you would manage those kidneys, it would be helpful. My final question concerns your postoperative follow-up, which appears to be mostly by ultrasound. Do you routinely, at any time, recheck your kidneys with a renal scan to see that they are showing good function and that they have good drainage? WH. Hendren (Bosron, MA): I agree completely with the authors’ thesis that when antenatal hydronephrosis is detected and persists postnatally it should be repaired in most cases. Surgeons who routinely operate for various neonatal problems can achieve excellent results performing pyeloplasties in young patients. Renal function at birth is about 20% of the adult normal value, which is acquired at about age 1 year. I believe that if moderate to severe obstruction is corrected early during the first year, when renal function is normally increasing, better function of that kidney results than if obstruction is relieved later. It is my firm belief that the conservative treatment of renal obstruction is to define the problem and fur it, using precise surgical technique. Dr Shaul, why do you think that there are many surgeons who believe that conservative treatment denotes long-term observation to document declining renal function or increasing hydronephrosis before proceeding with pyeloplasty in the small baby? A4. Dimler (Phoenix, AZ): Considering the prominence of the previous speaker, it is with great trepida-
tion that I make these remarks. I think the fact that you can operate on a newborn doesn’t necessarily mean that you should operate on a newborn. Philip Ransley (in England) and other authors (in a recent article in theJournal of Urology) have had the courage to follow up on some of these patients because they were perplexed by the increasing incidence of hydronephrosis with the frequent use of neonatal ultrasound and wondered why the condition had not been detected previously. The answer may be that in some of the patients the hydronephrosis disappears, and in following up on a lot of these patients, they have not demonstrated the predicted loss of renal function in patients who seem to have hydronephrosis. I don’t think the answer is known. It is certainly not known to me as to which patients should or should not be operated on, and I don’t think that this report has clearly answered that question. I think they have only said that they can operate safely in infants. D.B. Shaul (response): Thank you for your comments and questions. Dr Murphy, patients with severe hydronephrosis often have tubular dysfunction, and their nuclear renal scans can be misleading. The algorithm we follow is to rule out reflux and then document obstruction by noting progressive dilatation on serial ultrasounds or obtain a nuclear medicine diuretic renogram. The only exception would be the occasional patient with severe or massive hydronephrosis. Of course not all patients fit nicely into this algorithm, and for these we often obtain an intravenous pyelogram to visualize the anatomy of the UPJ preoperatively. Of the numerous infants we see with hydronephrosis, about 20% have sufficient evidence of obstruction to warrant surgery. The remainder are followed nonoperatively. Dr Hendren, I suspect that the reason people are not operating on these babies is that they accept the logic that “Before prenatal ultrasonography these patients were not detected until later in life, therefore it is harmless to wait until they are a little older.” That leads to Dr Dimler’s comment. I agree that just because you CCZYE operate on babies doesn’t mean you have to, but I also believe that we must modify our clinical practice to stay abreast of beneficial technological advances.
Is a Low-Risk
B. Shaul, James A. Cunningham, Phillip Lowe, Leonard Los Angeles, California
0 Early pyeloplasty for the treatment of congenital ureteropelvic junction obstruction to maximize nephron salvage is only justified if the potential hazards of operating on small infants are avoided. The records of all infants who underwent pyeloplasty by the authors over a 7-year period were analyzed. The outcome of surgery in the younger infant group (YIG; patients less than 2 months of age) was compared with that of the older infant group (OIG; patients more than 2 months of age). Preoperative evaluation in cases of mild or moderate hydronephrosis was directed toward ruling out a nonobstructed collecting system and included voiding cystourethrography and serial ultrasonography and/or dual isotope diuretic renography. Open pyeloplasty was performed if collecting systems had deteriorated or were demonstrated to be obstructed; it was also performed for severe cases of hydronephrosis. Postoperative assessment consisted of serial ultrasonography and/or nuclear imaging to confirm decompression and relief of obstruction. Thirty three pyeloplasties were performed in 31 patients in the YIG (two bilateral), and 33 were performed in 32 infants in the OIG (one bilateral). The only significant differences between the groups were as follows. Patients in the YIG were more likely to present in utero (87% Y 53%; P < .Ol), whereas those in the OIG were more likely to present with a urinary tract infection (44% Y 0%; P < .Ol). The YIG was more likely to have nephrostomy drainage postoperatively (67% 1145%; P < .Ol). In the overall series, there were five complications, all of which were postoperative infections requiring intravenous antibiotic therapy. Two occurred in the YIG and three in the OIG. One patient had a urinoma, which was drained percutaneously. During follow-up examination, 32 renal units in the YIG and 30 in the OIG were judged to be stable or improved. One kidney in each group was not salvaged after the pyeloplasty. Two patients in the OIG were lost to follow-up. No anesthetic complications or strictures occurred in either group. Good results of pyeloplasties performed in the infants in this series support early correction of ureteropelvic junction obstruction in infants. Copyright c 1994 by W.B. Saunders Company INDEX WORDS: Hydronephrosis, junction obstruction.
pyeloplasty;
ureteropelvic
I
NCREASES in the routine usage of prenatal ultrasonography has led to the diagnosis of hydronephrosis in utero in a large number of pregnancies. Ureteropelvic junction (UPJ) stenosis was found by prenatal ultrasound in 0.1% of pregnancies.’ The postnatal management of these dilated, potentially obstructed neonatal collecting systems is controversiai. McCrory et al were the first to demonstrate that relief of obstruction prevents further renal damage and, in a subgroup of infants, allowed for some functional recovery.’ Results of animal studies have also suggested that early relief of obstruction is JournalofPediafric Surgery, Vol 29, No 2 (February), 1994: pp 343-347
Procedure B. Skaist, and Brian E. Hardy
beneficial, even in cases of partial ureteral obstruction.3,4 This has led some investigators to encourage early pyeloplasty in cases of UPJ obstruction initially diagnosed antenatally. However, other investigators have been more selective and maintain that even severe hydronephrosis is not necessarily obstructive in nature.5-7 They have promoted observation with serial radiological evaluation for the management of unilateral neonatal hydronephrosis. The final functional outcome of these hydronephrotic, potentially obstructed kidneys remains to be seen. There is little doubt that early pyeloplasty in definite cases of UPJ obstruction will yield the best long-term functional result. In cases of significant hydronephrosis secondary to UPJ stenosis, it is also likely that pyeloplasty, when carried out with minimal complications, will provide the best ultimate outcome. This study was undertaken to determine whether there is a significant difference in postoperative results between young infants and older infants. MATERIALS
AND METHODS
The records of all infants undergoing pyeloplasty by the authors (between 1985 and 1992) were reviewed. Both inpatient and outpatient records were reviewed, and data were collected regarding patient demographics, associated anomalies, preoperative evaluation, hospital course, postoperative evaluation, and outcome. Patients were divided into two groups. based on age at the time of surgery. Patients undergoing pyeloplasty before 2 months of age formed the younger infant group (YIG), whereas those undergoing pyeloplasty at more than 2 months of age were in the older infant group (OIG). Preoperative evaluation was directed toward ruling out nonobstructive causes of hydronephrosis. Voiding cystourethrography (VCUG) was performed to rule out vesicoureteral reflux. All patients underwent confirmatory postnatal ultrasound. If severe or massive hydronephrosis was present with thinning of the renal parenchyma, pyeloplasty was performed. However. if mild or moderate hydronephrosis was present, serial ultrasonography was performed 3 to 6 weeks later. If deterioration was noted, the patient underwent pyeloplasty. If there was no change, dual isotope diuretic renography was performed to determine whether obstruction was present. Agents used included technetium 99m-diethylenetriaminepentaacetic acid (Y’hnT~-DTPA) and iodine 131 orthoiodo Hippuran (i3iI-OIH). Furosemide (0.5 mgikg) was administered 16 minutes after injection of the tracer.
From the Division of Pediatric Urology, Children’s Hospital, Los Angeles, CA. Presented at the 24th Annual Meeting of the American Pediatric Surgical Association, Hilton Head, South Carolina, May 15-18, 1993. Address reprint requests to Brian E. Hardy, MD. 1300 North Vermont Ave. Suite No. 708, Los Angeles. CA 90027. Copyright 0 I994 by KB. Saunders Cornpan? 0022-3468/2902-0040$03.0010 343
SHAUL ET AL
RESULTS
Fig 1. The stent is created by cutting extra side holes in a pediatric feeding tube. It is important to position the catheter so that all the side holes are in the pelvis or proximal ureter.
Patients treated in the last 5 years of the study period had diuretic-induced isotope washout curves constructed. In patients with good function (renal isotope excretion), the time to wash out half the isotope was measured and recorded as the tl/2. t’/r values of greater than 30 indicated obstruction. Cases of obstruction demonstrated by this method also underwent pyeloplasty. Patients whose kidneys showed improvement on serial ultrasonography, partial obstruction on nuclear medicine studies, or no obstruction did not undergo surgery and are not included in this series. In all patients the kidney was exposed via an extraperitoneal flank approach. The type of reconstruction used was the AndersonHynes dismembered pyeloplasty. Resection of redundant pelvis was performed only in cases of massive dilatation. Drainage of the bladder was routine for 2 to 4 days postoperatively. Patients were more likely to have nephrostomy drainage and/or stent placement when there was perirenal inflammation, recent infection, a solitary functioning kidney, or massive dilatation of the pelvis with calyces that were likely to be filled with debris. In the majority of patients a nephrostomy tube stent combination was fashioned using a 5F feeding catheter with multiple extra side holes (Fig 1). The catheter was placed such that the tip lay beyond the anastomosis in the proximal ureter, and the extra side holes lay within the pelvis of the kidney. The tube was brought out through the inferiormost calix, and a purse-string suture was placed in the renal capsule surrounding the tube to prevent extravasation around the tube. The tube was left plugged, but was unplugged if a nephrostomy was needed. Penrose drainage of the perirenal space was routinely employed, with the penrose brought out through a separate stab incision. Antibiotic prophylaxis was continued in the postoperative period for an average of 2 weeks. Postoperative evaluation consisted of serial ultrasonography. Progressive decrease in the amount of dilatation of the renal pelvis was equated with successful relief of obstruction. If sequential ultrasonography showed hydronephrosis or if there were concerns about the functional status of the kidney, the patients also underwent dual isotope diuretic renography to confirm that obstruction had been relieved. When possible, follow-up examinations were performed yearly, for up to 4 years postoperatively, to detect recurrence of UPJ obstruction or development of contralateral UPJ obstruction. Statistical analysis was performed using a statistical program (Stat View, Calabasas, CA). Normally distributed continuous variables were analyzed using the Student’s t test for unpaired data. Other continuous variables were analyzed using the MannWhitney U test. The ,$ test was used to analyze noncontinuous variables.
Thirty three pyeloplasties were performed in 31 infants in the YIG, and 33 were performed in 32 infants in the OIG. Three infants underwent bilateral pyeloplasty, two of which were performed under the same anesthetic (Table 1). There was male predominance in both groups. The left side was involved approximately twice as often as the right side. The mean age of YIG patients was 36 + 16 days, and the mean age of OIG patients was 160 + 91 days. There was a significant difference in the mode of presentation between the groups (Table 2). Patients in the YIG were more likely to present as the result of a prenatal ultrasound finding than were those in the OIG (87% v 53%; P < .Ol). Patients in the OIG were more likely to present with a urinary tract infection (44% v 0%; P < .Ol). In the total series, five patients presented with an abdominal mass. One patient with a contralateral multicystic dysplastic kidney presented with acute renal failure secondary to the UPJ obstruction. Associated anomalies were most likely to have involved the genitourinary system (Table 3). Five infants had a contralateral multicystic dysplastic kidney (11%). All these patients were in the YIG when early decompression of their contralateral obstruction was performed. Vesicoureteral reflux occurred in four patients (7%). In two instances, the reflux was into the obstructed unit, and in two it was into the contralateral kidney. Four infants had contralateral hydronephrosis. These cases were evaluated with dual isotope diuretic renography. Two were partially obstructed, and two were not obstructed. In all four the hydronephrosis resolved without operative treatment. Studies used in the preoperative evaluation of the patients are shown in Table 4. Voiding cystourethrography (VCUG) was performed in 89% of the patients preoperatively. Patients who did not have preoperative VCUG were those for whom the diagnosis of obstruction was clearly established by other methods of investigation. With the exception of one patient for whom the diagnosis was made preoperatively from an incidental finding on a computed tomography scan, all patients underwent ultrasonography. Diuretic renography was performed preoperatively to quantitate Table 1. Patient Demographics Young
Age No. of patients No. of pyeloplasties No. of bilateral pyeloplasties
Infants
Older Infants
<2mo 31
>2mo
33
33
2
32 1
Sex (male:femalel
22:9
23:9
Side (right:left)
12:17
8:23
345
INFANT PYELOPLASTY
Table 4. Preooerative Evaluation
Table 2. Mode of Presentation Young
Older
Infants
Young Infants
Infants
In = 31)
(n = 3.2)
(n = 311
Older
Infants
In = 321
27
17*
Voiding cystourethrography
28
28
Urinan/ tract infection
0
14’
Renal ultrasound
30
32
Mass
2
3
Acute renal failure
1
0
Prenatal ultrasound
t% washout
NOTE. Patients may have more than one mode of presentation. *P i
21
23
70.4 + 33
64 r 28*
15
18
0
1
isotope renal scan
.Ol.
Intravenous
(min) pyelography
Whitaker test *P is not significant.
the degree of obstruction in 44 patients (70%). Values for t’/z were 70.4 -+ 33 minutes in the YIG and 64 +- 28 minutes in the OIG. These values were not significantly different. Intravenous pyelography was performed on 33 patients, and one patient had a Whitaker test preoperatively. Five patients had lower urinary tract infections postoperatively (two in the YIG, three in the OIG). All resolved with antibiotic therapy. One patient had a urinoma, which was drained percutaneously and did not recur. Other postoperative results are shown in Table 5. Younger infants were more likely to have their anastomosis protected by nephrostomy drainage and/or stent placement than were older infants. The length of postoperative stay was not different between the groups. Patients in the YIG had significantly longer follow-up than did those in the OIG (23 & 21 months v 12 + 12 months). Seventy-three percent of the patients in each group had follow-up for at least 3 months. Two patients in the OIG did not return for postoperative studies; the 64 remaining renal units had between one and five postoperative ultrasounds, with an average of 2.5 postoperative ultrasounds per patient in each group. Improvement of hydronephrosis was noted for 30 renal units in the YIG and 29 units in the OIG. Three patients did not return for additional follow-up studies after an early postoperative ultrasonography demonstrated no change in the hydronephrosis. Although many cases with this finding had improvement subsequently, the ultimate outcome of these three patients is not known. Nine patients had postoperative isotope scans; there was Table 3. Associated Genitourinary
Vesicoureteral
Older
Infants
Table 5. Postoperative
In = 31)
h = 32)
5
0
1
1
Postoperative
Contralateral
0
2
Follow-up
0
2
1
1
Crossed ectopia
1
0
Improved
drainage
stay (d)
(mo)
No follow-up
nonobstructive
hydronephrosis
(n = 33)
Nephrostomy/stent
partial UPJ
obstruction
Results
Young Infants
reflux
lpsilateral
Contralateral
Changes in renal function after relief of urinary tract obstruction in pediatric patients was initially studied by McCrory et al.’ Creatinine clearance, urinary concentrating ability, and excretion of an acid load were studied in 15 patients before and after treatment of urinary tract obstruction. These investigators noted that infants have the greatest capacity for improvement after relief of obstruction. It is well known that renal function continues to mature after birth.6 Elimination of potentially harmful obstruction from these maturing nephrons makes sense intuitively. The benefit of early decompression has been applied to large groups of neonates and infants, with good results.8-1”Simultaneously, experimental analyses were performed in different animal models.3,4 A series of experiments in a neonatal guinea pig model further elucidated the effects of permanent and temporary partial unilateral ureteral obstruction on renal function. It was noted that even though the intraureteral pressures were returned to normal after relief of obstruction, renal functional impairment persisted into adulthood.14 Even though these carefully performed studies appear to show the harmful effects of untreated unilateral ureteral obstruction on renal function, the treatment of neonatal hydronephrosis in humans has remained controversial.
dysplastic
Kidney (contralateral)
Contralateral
DISCUSSION
Anomalies
Young Infants
Multicystic
normal function in three, adequate function (but a small kidney) in four, and no function on the involved side in two. The latter six patients had markedly dilated collecting systems preoperatively.
studies
Older
Infants
(n = 33)
22
15*
7.7 f 6.1
5.8 f 5.0
23 f 21
12 f 12*
0
2
30
29
Stable
2
1
Not salvaged
1
1
lP < .Ol.
346
SHAUL ET AL
The appropriate treatment for neonates diagnosed in utero as having hydronephrosis secondary to anomalous UPJ was addressed by Homsy et aL5 Twenty-one of 60 renal units were believed to be partially obstructed or nonobstructed, despite being hydronephrotic. Because deterioration occurred in only 41% of patients not treated operatively, the investigators recommended a period of observation for all patients in whom obstruction could not be clearly demonstrated. Similar nonoperative approaches have been adopted by other investigators.6,7 A consistent conclusion among these investigators is that a definite diagnosis of obstruction is difficult to make. For example, the Whitaker test was found to be unreliable in the diagnosis of neonatal ureteral obstruction.’ In another series, all the standard modalities used to detect obstruction were unreliable when applied to neonates. It was suggested that contralatera1 compensatory hypertrophy is the most sensitive indicator of obstruction.6 The combination of evidence demonstrating the harmful effects of untreated obstruction on neonatal kidneys and the difficulty one faces in proving obstruction makes the approach of observing potentially obstructed neonatal kidneys worrisome. If UPJ obstruction can be safely relieved in these young infants, the potential damage that occurs during the observation period may be avoided. The safety with which pyeloplasty can be performed in neonates and young infants has been demonstrated.8-1”$‘5 Sixty-six pyeloplasties were performed in 63 infants in our series. Roughly half these procedures were performed in infants less than 2 months of age. Despite the young age and small size of the infants, there was no appreciable difference in outcome between this group and the OIG. There were no anesthetic complications and no deaths in the series. Follow-up was inadequate in five cases. Of the remain-
ing kidneys, postoperative ultrasonography demonstrated improvement in all cases except some that had a severe degree of dilatation preoperatively. Performance of pyeloplasty in the YIG did not prolong the hospital stay. Younger patients were more likely to have their repairs protected by some combination of nephrostomy and/or stent placement. These tubes were always removed 1 to 2 weeks after surgery. This type of drainage is considered to be a safeguard rather than a necessity. The fact that all tubes were able to be removed within 2 weeks confirms that significant anastomotic obstruction did not occur. There is no evidence that placement of these tubes promotes anastomotic obstruction, as has been suggested.6 The longer follow-up in the YIG is a reflection of parental compliance. Early postnatal treatment was sought for more than 80% of infants in the YIG. Even though nearly 50% of OIG cases were diagnosed in utero, most parents did not seek treatment for their infants until several months after delivery. This pattern continued postoperatively. It is interesting to note that 44% of the patients in the OIG presented with a lower urinary tract infection. The same finding has been noted by others.711” The combination of urinary stasis in the dilated upper tract as well as infection in the lower tract creates the possibility of a serious urinary infection. For this reason, urinary prophylaxis has been recommended until dilatation is relieved.‘” Unfortunately, the definitive diagnosis of UPJ obstruction is often difficult to make. The relief of obstruction by resection of the UPJ and performance of a dismembered pyeloplasty has now been safely achieved for large groups of infants. We believe that prolonged observation of those kidneys, which are likely to be obstructed, is currently unjustified.
REFERENCES 1. Grignon A, Filiatrault junction stenosis: Antenatal investigation, and follow-up.
D. Homsy Y, et al: Ureteropelvic ultrasonographic diagnosis, postnatal Radiology 160:649-651. 1986
2. McCrory WW, Shibuya M, Leumann E, et al: Studies of renal function in children with chronic hydronephrosis. Pediatr Clin North Am l&445-465, 1971 3. Chevalier RL, Kaiser DL: Chronic partial ureteral obstruction in the neonatal guinea pig. Influence of uninephrectomy on growth and hemodynamics. Pediatr Res l&1266-1271, 1984 4. Taki M, Goldsmith DI, Spitzer A, et al: Impact of age on effects of ureteral obstruction on renal function. Kidney Int 24:602-609, 1983 5. Homsy YL, Williot P, Danais S: Transitional neonatal hydronephrosis: Fact or fantasy? J Ural 136:339-341, 1986 6. Koff SA, Campbell K: Nonoperative management of unilateral neonatal hydronephrosis. J Urol148:525-531,1992 7. Dhillon
HK,
Duffy
PG,
Gordon
I. et al: A ramdomised
clinical trial in infants with prenatally diagnosed unilateral hydronephrosis. Presented at the meeting of the Section on Urology of the American Academy of Pediatrics. San Francisco. CA. October 12, 1992 8. Bevani B, Belman BA: Ureteropelvic newborns and infants. J Urol 128:770-773.
junction 1982
obstruction
9. Roth PR. Gonzales ET: Management of ureteropelvic tion obstruction in infants. J Ural 129:108-110, 1983
in
junc-
10. King LR, Coughlin PWF, Bloch EC, et al: The case for immediate pyeloplasty in the neonate with ureteropelvic junction obstruction. J Urol 132:725-728. 1984 11. Murphy JP, Holder TM, Ashcraft KW, et al: Ureteropelvic junction obstruction in the newborn. J Pediatr Surg 19:642-648, 1984 12. Flake AW, Harrison MR, Save L, et al: Ureteropelvic junction obstruction in the fetus. J Pediatr Surg 21:1058-1063, 1986
INFANT PYELOPLASTY
347
13. Wolpert JJ, Woodard JR, Parrott TS: Pyeloplasty in the young infant. J Urol 142573-575. 1989 14. Chevalier RL: Renal response to ureteral obstruction in early development. Nephron 56:113-l 17, 1990
15. Bernstein GT, Mandell J. Lebowitz RK, et al: Ureteropelvic junction obstruction in the neonate. J Ural 140:1216-1221.1988 16. Koff SA: Presented at the Los Angeles Urological Society, September 1992
Discussion In reviewing your manuscript, Dr Sham, several questions come to mind. First, it appears that you selected your patients on the basis of severe dilation, and the ones that were the most hydronephrotic were the ones that were operated on. Also, patients whose renal scan showed decreased function or delayed drainage were operated on too. It was a little unclear as to the number of patients that were observed nonoperatively, and the method of follow up. If you could clarify that for us, it would help. Second, do the conclusions of your report indicate that you are going to treat these kidneys differently in the future? Are you going to be more aggressive with operating on them, or are you going to manage them in the same way? If you could give us an idea of your algorithm on how you would manage those kidneys, it would be helpful. My final question concerns your postoperative follow-up, which appears to be mostly by ultrasound. Do you routinely, at any time, recheck your kidneys with a renal scan to see that they are showing good function and that they have good drainage? WH. Hendren (Bosron, MA): I agree completely with the authors’ thesis that when antenatal hydronephrosis is detected and persists postnatally it should be repaired in most cases. Surgeons who routinely operate for various neonatal problems can achieve excellent results performing pyeloplasties in young patients. Renal function at birth is about 20% of the adult normal value, which is acquired at about age 1 year. I believe that if moderate to severe obstruction is corrected early during the first year, when renal function is normally increasing, better function of that kidney results than if obstruction is relieved later. It is my firm belief that the conservative treatment of renal obstruction is to define the problem and fur it, using precise surgical technique. Dr Shaul, why do you think that there are many surgeons who believe that conservative treatment denotes long-term observation to document declining renal function or increasing hydronephrosis before proceeding with pyeloplasty in the small baby? A4. Dimler (Phoenix, AZ): Considering the prominence of the previous speaker, it is with great trepida-
tion that I make these remarks. I think the fact that you can operate on a newborn doesn’t necessarily mean that you should operate on a newborn. Philip Ransley (in England) and other authors (in a recent article in theJournal of Urology) have had the courage to follow up on some of these patients because they were perplexed by the increasing incidence of hydronephrosis with the frequent use of neonatal ultrasound and wondered why the condition had not been detected previously. The answer may be that in some of the patients the hydronephrosis disappears, and in following up on a lot of these patients, they have not demonstrated the predicted loss of renal function in patients who seem to have hydronephrosis. I don’t think the answer is known. It is certainly not known to me as to which patients should or should not be operated on, and I don’t think that this report has clearly answered that question. I think they have only said that they can operate safely in infants. D.B. Shaul (response): Thank you for your comments and questions. Dr Murphy, patients with severe hydronephrosis often have tubular dysfunction, and their nuclear renal scans can be misleading. The algorithm we follow is to rule out reflux and then document obstruction by noting progressive dilatation on serial ultrasounds or obtain a nuclear medicine diuretic renogram. The only exception would be the occasional patient with severe or massive hydronephrosis. Of course not all patients fit nicely into this algorithm, and for these we often obtain an intravenous pyelogram to visualize the anatomy of the UPJ preoperatively. Of the numerous infants we see with hydronephrosis, about 20% have sufficient evidence of obstruction to warrant surgery. The remainder are followed nonoperatively. Dr Hendren, I suspect that the reason people are not operating on these babies is that they accept the logic that “Before prenatal ultrasonography these patients were not detected until later in life, therefore it is harmless to wait until they are a little older.” That leads to Dr Dimler’s comment. I agree that just because you CCZYE operate on babies doesn’t mean you have to, but I also believe that we must modify our clinical practice to stay abreast of beneficial technological advances.