Pediatric Nephrology

Pediatr Nephrol (1995) 9:503-509 9 IPNA 1995

Urological review Neonatal hydronephrosis- the controversy and the management Benjamin M. Tripp and Yves L. Homsy Division of Pediatric Urology,Departmentof Surgery,Montreal Children's Hospital, McGill University,Montreal, Quebec, Canada Received November 3, 1994; received in revised form December 7, 1994; acceptedDecember 20, 1994

Abstract. Neonatal hydronephrosis is being detected with increasing frequency. The majority of these cases have a tendency to resolve during infancy. Hydronephrosis is an anatomical entity that is not synonymous with obstruction. Review of the pathoembryology, the pathophysiology, the diagnostic techniques currently used and the natural history of hydronephrosis is given. The management and the controversies involved are discussed.

Key words: Hydronephrosis - Kidney diseases - Prenatal diagnosis - Ultrasonography - Nuclear renography

Introduction Hydronephrosis (HN) is an anatomical entity that is defined as an enlargement of the capacity of the collecting system of the kidney, calices and pelvis. Hydroureteronephrosis (HUN) includes the dilatation of the ureter. It has become clear that HN and HUN are not synonymous with obstruction. Currently, obstruction is defined clinically as a restriction to urinary ou~ow that, left untreated, will cause progressive renal deterioration [1]. This definition is accurate, measurable and useful in dealing with obstruction that may require surgical correction [1]. Unfortunately, current techniques cannot reliably diagnose obstruction without including an observation period and therefore, one must wait for the development of measurable progressive renal injury or the development of tests that are more sensitive in picking out kidneys that are being damaged [1]. Today, with the introduction and explosion of the number of ultrasonographic examinations being performed prenatally, the impact on early detection of congenital anomalies has been staggering.

Correspondence to: Y. L. Homsy, Montreal Children's Hospital, 2300 Tupper Street, Montreal, Quebec, H3H 1P3, Canada

HN is being detected with increasing frequency [2-4], in as many as 1.4% of fetuses, persisting postnatally in half of these cases [5]. The patient is now presenting to the urologist or nephrologist before he or she is even born, with a presumptive diagnosis rather than a symptom. Evaluating and managing HN are issues that are evolving rapidly and hence generating enormous controversy. The significance of HN in neonates and infants is often difficult to define. This is true because one cannot distinguish whether HN is due to an ongoing obstruction that will invasively cause further kidney damage or whether it is a sequel of an obstructive event that both occurred and resolved antenatally. In cases where the etiology is recognizable and understood (e. g., posterior urethral valves, ureterocele), conventional treatment to re-establish optimal urinary drainage is accepted. However, such cases occur rather infrequently. In the differential diagnosis of HN, Homsy et al. [6] list the most common causes of HN in neonates in order of frequency, ureteropelvic junction (UP J) anomalies being the most common etiology. Found in decreasing frequency are ureterovesical junction (UVJ) anomalies, vesicoureteral (VU) reflux, multicystic kidney and posterior urethral valves (Fig. 1). Other causes include obstructive and nonobstructive megaureter, ureterocele, nenrogenic bladder, prune-belly syndrome and urethral atresia. Neonatal HN evokes several basic questions without any consensus about the answers. What is the functional significance of HN? When do we operate and when do we choose to observe infants with HN? If we decide to operate, when is the optimal time? How do we know which infants with HN will deteriorate and which will stabilize or improve? What is the optimal follow-up of infants with HN? These are the clinical dilemmas that face most pediatric nephrologists and urologists. Understanding the pathophysiology, the pathoembryology, the diagnostic techniques currently in usage and the natural history of HN is essential if one is to manage postnatal HN adequately. This article will review the current state of knowledge and will outline a practical approach to neonatal HN.

504

UPJ obstruction 64.0%

6 ~ PUV 2.1% MCK 4.3% n1 3 . 4 ~ % UVJ obstructio

.

. . . . . . . . . . ~!iii~,. ~C'.'2~'~F ~~~Defunct ~'~ ~%

~U

o k dney 32%

reflux 9.1%

Fig. 1. Etiology of hydronephrosis. UPJ, Ureteropelvicjunction; UVJ, ureterovesicaljunction; VU, vesicoureteral;MCK, multicystic kidney; PUV, posterior urethral valves

Pathophysiology In recent years, critical appraisal of current knowledge of the pathophysiology of renal obstruction has resulted in more insight into the complex factors influencing the natural history of this anomaly. Obstructive uropathy is currently being perceived as a process with characteristic functional, biochemical histological and anatomical changes. Previously, the consequences of obstruction were classically believed to occur solely via pressure-induced mechanisms, now it can be understood possibly as a primary hemodynamic event that results in a form of ischemic injury. Inroads to understanding the pathophysiology of obstruction have recently been made and are reorienting our understanding of this condition. Obstructive uropathy is now being understood primarily as renal disease resulting from diminished urinary drainage, otherwise known as obstructive nephropathy [7]. The classic concept of pressure-related mechanisms of renal impairment is now being revised to one of a hemodynamic response to obstruction possibly causing ischemic injury [8]. Renal damage may be produced by vasoconstrictive ischemia [9]. Functionally there exists a spectrum of obstruction from complete to partial to nonobstructive. HN, therefore, can refer to any of these entities. Partial obstruction by its very nature is variable and is difficult to measure and so are the effects and responses of partial obstruction. Partial ureteric obstruction has reliably been produced in many animal models [9-13]. Reviewing three excellent and separate models that study the well-documented renal hemodynamics of partial obstruction, one could conclude that ureteral obstruction primarily causes preglomerular arteriolar vasoconstriction and that this can mediate the ischemic renal damage seen so often in obstructive uropathy in neonates and infants. A congenital HN model in rats was introduced by Lewy [14]. Renal function of the hydronephrotic kidney was half that of the contralateral side. The renin-angiotensin system was found to be a direct mediator of this reduced renal function [14]. In a second rat model of partial ureteral

obstruction via insertion of internal ureteral stents, HN and reduced renal function were seen [11]. This renal damage can be mediated via preglomerular vasoconstriction. In a third model proposed by Chevalier [12], ischemic changes were caused by vasoconstriction, also shown to be dependent on the rehin-angiotensin system. These models show that partial obstruction is a complex process causing a cascade of interactions, including the changes in the major tubular enzymes responsible for secretion and reabsorption [7] and the synthesis of vasoactive compounds such as angiotensin II, thromboxane A2, nitric oxide and eicosanoids which can mediate preglomerular vasoconstriction [8] and renal damage indirectly or mediate renal damage directly.

Pathoembryology Using the ureteral bud as the reference point, in the 5th week of gestation (as the embryo elongates and the hindend forms), the ureteral bud, hollow at first, arises dorsally from the mesonephric or wolffian duct. It penetrates the metanephric blastema inducing the formation of the ureter, pelvis, calices and collecting ducts from the bud and the entire excretory system from the blastemal cap. Ascent of the kidney occurs between the 4th and 6th week, and nephrogenesis takes place at 8 weeks and is complete by 34 weeks. Urine is first formed at 10 weeks of gestation and the fetal kidneys become the source of 90% of amniotic fluid at 16-20 weeks. Fetal kidneys and HN can be first detected by ultrasound at 15 weeks' gestation but HN becomes obvious at 20 weeks, the time at which the kidney is surrounded by fat and the internal renal structures appear distinct [15]. Transient anatomical obstruction is proposed as a major mechanism of antenatal HN. Causes include persistence of a "Chwalla's-like" ureterovesical membrane [16], natural folds and kinks in the ureter as described by Ostling [17] and failure of the recanalization of the ureter as described by Ruano-Gil et al. [18]. Ruano-Gil et al. [18] state that during the normal ascent of the ureter, the tubular lumen solidifies and at day 40 of pregnancy recanalization begins. Starting in the mid-portion of the ureter, it progresses towards both the UVJ and UPJ. Therefore, both the UPJ and the UVJ are the last sections of the ureter to be recanalized. Umbilical arteries causing extrinsic pressure on the ureters have also been suggested as a possible cause of transient obstruction [19].

Diagnostic techniques in neonatal HN The majority of neonates with HN are detected antenatatly, most of whom are asymptomatic, have a normal physical examination, a normal creatinine and a negative urinalysis. These neonates have no indication of any urological or renal abnormality. Over the past decade, it has become apparent that dilatation of the upper urinary tract does not necessarily mean that obstruction is present [6, 20]. In addition, it is unknown if partial obstruction leads to shortor long-term functional renal deterioration. The controversy

505 of evaluating, managing and following hydronephrotic kidneys hinges on the assessment of renal function over a period of time. We will be discussing the advantages and pitfalls of the different ways of evaluating neonatal kidneys. Diagnostic techniques in HN are constantly evolving and it is imperative to view HN as a disease for which we are measuring only certain consequences on the kidney that are heavily dependent on the nature of the test. There is no single test to accurately determine all of the neonatal kidney's functions.

Ultrasonography Ultrasound has become widely used in diagnosing antenatal HN and is the mainstay of screening, although it has recently found a role in follow-up [21]. Ultrasonography provides an excellent morphological assessment of perinatal kidneys. A classification of in-utero HN based on more than 34,000 maternal ultrasounds performed after the 20th week of gestation was developed by Grignon et al. [5]. HN ranged from detectable renal pelvic dilatation (grade I) to pelvic dilatation greater than 1 cm (grade II) to three degrees of progressive caliectasis (grade III-V). Ultrasonography predicted that the group of mildly dilated renal units, the most common finding on ultrasound, were most likely to resolve spontaneously soon after birth. The degree of caliectasis and pyelectasis was related to the degree of cortical thinning. Maizels et al. [22] described a postnatal classification of HN and HUN which is commonly used and consistent among pediatric radiologists and urologists and has been adopted by the Society for Fetal Urology. Problems associated with the diagnostic accuracy of the ultrasonogram are mainly secondary to operator dependency. Also, ultrasound should only be performed after the first few days of life because of the physiological dehydration which may occur causing neonates with HN to be missed. When HN is moderate or severe, the initial ultrasound can have important value in predicting functional renal impairment. The use of Doppler ultrasonography to measure the resistive index (RI = peak systolic velocity - end diastolic velocity/peak systolic velocity) has been variably reliable in diagnosing obstruction in children in comparison to its usage in adults [23]. Keller et al. [24] used the resistive index ratio (RIR = obstructed RI/unobstructed RI) to determine if obstruction was present. The correlation was better but still lacks specificity. Diuretic Doppler ultrasonography shows promise in evaluating HN, and combining the RIR and the diuretic Doppler ultrasound might be a valuable new diagnostic test [25]. In unilateral hydronephrosis, ultrasound is now being proposed to accurately measure compensatory changes in the opposite non-hydronephrotic kidney. Koff et al. [21] have proposed plotting serial measurements on a renal growth chart to identify infants with obstruction as evidence by accelerated contralateral renal growth (compensatory hypertrophy). This phenomenon has also been noted in utero [26].

Voiding cystourethrography and intravenous urography Voiding cystourethrography must be performed in all cases of a dilated urinary tract in the neonate. Residual reflux may be responsible for HN and/or HUN, and is associated in 15% of patients with UPJ "obstruction" [27] und UVJ "obstruction" [28]. Vesical or infravesical obstruction must also be ruled out even if upper tract obstruction is presumed. Intravenous urography (IVU) is the traditional method for evaluating HN but has a limited role today in neonates and infants. IVU should be used in older infants and children and is still the study of choice when one wants to visualize ureteral and renal anatomy. However, it does have major drawbacks in the neonate with a dilated upper tract. These include physiological factors that are classically described in neonates, such as poor concentrating ability, poor proximal tubule reabsorption and a low glomerular filtration rate (GFR). In addition, overlying abdominal gas patterns often obscure the kidneys and ureters.

Whitaker pressure flow study The Whitaker test is a pressure perfusion-study measuring the response of the renal pelvis to distension and is based upon the principle that the resistance is equal to the pressure divided by the flow. In older children and adults, this test has been reported to be accurate in about 80% of cases and equivocal in 20%. The correlation with diuretic renography varies, ranging from 40% to 60% [29]. However, in neonates it has significant limitations. Foremost is its invasive nature involving a percutaneous kidney puncture, bladder catheterization and the requirement of some form of sedation. This test uses flows that are nonphysiological in the neonate and infant. It also assumes that obstruction produces a constant restriction to outflow and not all obstructions are constant. Other pitfalls include varying techniques and normal values of perfusion rates and pressures in neonates that have not been validated. A last consideration is that loss of function cannot be monitored by the Whitaker test. For these reasons, the Whitaker test has fallen into disfavor and is less frequently used in the pediatric population.

Nuclear renography Nuclear renography with and without diuresis is a recent but established technique with well-defined capabilities and limitations. The usefulness of nuclear renography can be fully reached when certain concepts are used in its interpretation. Firstly, one must compare the dilated kidney with the nondilated kidney. Secondly, differential and absolute renal function must be determined and lastly, these determinations must be followed over time. Renal scintigraphy is the most commonly used diagnostic modality to assess and follow renal function and urinary tract drainage. 99mTechnetium-diethylenetriaminepentaacetic acid (99mTc-DTPA) is rapidly cleared by glomerular filtration and can be utilized to estimate

506 relative renal function. 99mTechnetium-dimercaptosuccinic acid (99mTc-DMSA) renal scintigraphy is also important in HN to measure the relative amount of functioning renal mass. DMSA remains bound to the cortical proximal renal tubules with minimal excretion, therefore technically it is easier to estimate renal function in HN with the DMSA scan than DTPA. This can be important in moderate to severe HN. 99mTechnetium-mercaptoacetyltriglycine is a newer tubular function agent that gives a superior image and the estimated renal plasma flow indirectly [30]. Accurate evaluation of renal function in absolute and relative parameters are quite important. Unfortunately, to date, the accuracy of gamma-camera techniques to measure GFR or estimated renal plasma flow has been limited in children and, furthermore, the multiple blood samples required limit the technique even further. Lambert et al. [31] described a technique, the "external skull counting method," in children involving a 99mTcDTPA renal scan and only a single blood sample that can reproduce the accuracy of multiple blood sampling. This technique determines absolute GFR and the differential GFR. In a prospective study of 99 patients, 0 . 5 - 1 7 years of age (mean 10.5 years), where 18 patients were less than 3 years of age, there was a correlation of t2 = 0.988, confirming reliability in calculating GFR against the standard four plasma sample monoexponential technique. This new adaptation of the renal scan allows a better evaluation of renal function and provides an additional parameter to describe and analyze kidney status in neonates and children which was not possible previously without multiple blood samplings.

Diuretic renography The diuretic renogram is a type of a renal scan that is primarily used to separate nonobstructive HN from obstructive HN. The diuretic renogram is a physiological study that assesses the kidney's ability to respond to a diuretic-induced volume challenge. The underlying basis of diuretic renography is that in the absence of obstruction, the radionuclide that has accumulated in the collecting system will be "washed out" by the diuretic effect of the furosemide which will continue to fill the collecting system with urine that does not contain any significant radionuclide. In the presence of obstruction, this washout will not occur because radionuclide cannot be physically cleared from the collecting system. Presently, there exists a large proportion of children, particularly neonates and infants, where there is an equivocal drainage time as determined by renal scan, and obstructive HN cannot always be differentiated from nonobstructive HN [1, 32]. The diuretic renogram is simple and safe and it provides significant functional information in a minimally invasive manner. Kass and Maid [33] described a quantitative method of measuring the clearance time required for 50% of the radionuclide to exit the collecting system and called it T1/2. Results were usually expressed in three categories using the T m or the half-time. A T 1/2 of < 10 min meant no obstruction, a T 1/2 of 1 0 - 2 0 rain was indeterminate and a T 1/2 > 2 0 rain meant obstruction was present. It soon became

obvious that the results reported showed some inconsistencies. If one is to truly understand the subtleties of diuretic renography, one has to review its pitfalls. It is now evident that diuretic renography is not that simple.

Pitfalls of nuclear renography. There are many factors that can significantly affect the results of the diuretic renogram. Results can be heavily operator-dependent and the techniques and standards of the renogram can vary from one institution to another. The controllable factors that relate to the technique include: (1) hydration status, (2) bladder fullness, (3) diuretic dose, (4) timing of diuretic administration, (5) method utilized for calculating the T m, (6) region of interest, (7) patient position and (8) use of a standard radiopharmaceutical. In an attempt to standardize the controllable factors that can determine the outcome of diuretic renography and so that the results could be interpreted and compared between institutions, a common methodology was adopted and a standardized diuretic renogram was established, the "welltempered renogram" [34, 35]. The "well-tempered renogram" is a technique that involves bladder catheterization when indicated, hydration by intravenously administered fluids, DTPA and furosemide injection in standard doses and at standardized times and delay until the neonate is 1 month of age. The major remaining pitfalls are the uncontrollable factors that influence the outcome of the test and are related to the physiological aspects of the diuretic renogram. The first and most important factors are renal function and the ability to respond to a diuretic. This is very important in cases where renal function is so marginal that the diuretic does not cause increased urine formation. In neonates, there are unique physiological changes that are ongoing and the validity of the obstructive pattern in the diuretic renal scan has been questioned by Koff et al. [36] and Ransley et al. [39]. There is often an immature and blunted renal response to furosemide and an inability to concentrate the radionuclide load. There is also a significant increase in the renal function and GFR, the fastest rate of increase occurring within the first 12 weeks of life and reaching a plateau at 2 years of age. Koff et al. [37] further state that the nonaffected kidney T 1/2 provides a qualitative assessment of global renal function and can define the level of renal functional maturity. However, the controversy on the validity of carrying out a renal scan during the 1st month of life still remains. A second source of error is the dynamics involved due to the variable compliance of the urinary system. If the compliance of the collecting system above the level of obstruction is low due to hypertrophied muscle fibers, intrapelvic pressure may increase sufficiently to overcome the obstruction and the T 1/2 may be normal in the face of obstruction. If the system is highly compliant, distension without elevation of intrapelvic pressures may occur, and the T v2 may be prolonged in the absence of obstruction. The third important source of error is the "volume dilution effect" or "mixing chamber" effect. When the dilatation is so great in relation to the diuresis produced, the system cannot be flushed of the radionuclide. In other words, the exchange rate of nonradioactive urine into the

507 large volume chamber will produce a prolonged T 1/2. It is important that the scan be interpreted in the overall clinical setting and determination of restricted flow should not be based only on the TI/2 but must take into consideration the shape of the washout curve and the appearance of the sequentially obtained images. As well, it is important to note the blood flow of the kidney, renal size, time of peak renal activity, which is normally 3 - 5 min post injection, and the absolute time required for the radionuclide to be excreted by the kidneys. Determining clinical obstruction as previously defined is difficult because current diagnostic techniques fall short and do not provide an accurate and reliable evaluation of obstruction and possible loss of renal function with time. It was realized that many cases of delayed or absent drainage exhibited a transitional form of HN, before stabilizing or resolving spontaneously [6, 20]. Even more striking was the improvement of renal function after spontaneous improvement in drainage [6, 20].

Natural history of neonatal HN Clinical studies have recently investigated the natural history of prenatal, neonatal and infant HN. The evidence is clear that the majority of FIN is mild and most resolve spontaneously. Koff and Campbell [38] followed neonates with unilateral HN suggestive of UPJ obstruction and found that only 7% of patients had progression of HN or deterioration of function requiring pyeloplasty. Conservative management in 93% of patients resulted in no loss of renal function. Follow-up was determined by the relative function of the hydronephrotic kidney. Less than 20% required a repeat renogram every 2 weeks, < 3 0 % every month, < 4 0 % every 2 months and > 4 0 % every 3 months. In addition, the neonates that ultimately required pyeloplasty did not have any associated permanent loss of renal function. Homsy et al. [20] reported observations on 41 kidneys with HN secondary to a UPJ anomaly, the most common identifiable cause of fetal HN. They all had good function but demonstrated equivocal washout on diuretic renal scan; 88% (33 of 41 units) of these kidneys showed improved or stable renal function after a minimum of 1 year of followup. This spontaneous improvement was termed "transitional HN". Of 41 units, 8 (20%) showed evidence of deterioration of drainage and required pyeloplasty. Ransley et al. [39] reviewed a series of 100 neonates with UPJ obstruction; 77% maintained normal or improved function and 14% showed deterioration requiring pyeloplasty. Several authors have conservatively managed children with HN associated with UPJ anomalies [20, 38, 40]. They showed that as the drainage improved spontaneously, so did the function of the kidney. When renal function does deteriorate, it can do so rapidly, as observed by Homsy et al. [20]. Flashner et al. [41] observed a group of patients in whom the initial HN was not associated with obstruction but who subsequently developed obstruction. Follow-up averaged 1 year but deterioration was noted as late as 33 months. These observations led to the recommendation that

close follow-up of HN must be arranged until a new diagnostic modality is developed that can preselect kidneys at risk of deterioration. Similarly when UVJ obstruction causes HN and HUN, there is spontaneous resolution in the majority of cases. Liu et al. [42] reported 67 kidneys prenatally with primary obstructed megaureter. These children were initially treated conservatively and only 17% (11 of 67 units) had subsequent deterioration of renal function or urinary tract infection requiring surgery. The remainder showed stable or improved renal function and resolving dilatation upon follow-up.

Management of HN When I-IN is associated with VU reflux, nonobstructive megaureter, ureterocele, ectopic ureter, neurogenic bladder or prune-belly syndrome, management depends upon proper diagnosis, concurrent conditions and presence, location and severity of obstruction. Operative repair is indicated for symptomatic HN where there is urinary tract infection, pain or colic, failure to thrive, hematuria, calculi or hypertension. In bilateral HN, management is clear when infravesical obstruction is causing this entity. Surgery to decompress this obvious obstruction is needed. Adequate drainage is usually indicated when there is severe bilateral HN or significant bilateral depression of renal function. This will afford the best chance for the neonate to recover function. When overall function is normal and bilateral HN is mild and not due to infravesical obstruction, we follow these patients carefully as if they had unilateral HN. The majority of HN is unilateral and secondary to UPJ and UVJ anomalies. Treatment of HN in these cases is in a state of flux and is determined by the "comfort zone" of the individual urologist [43]. There are three main ways that have been suggested to determine this comfort zone. Firstly, the relative difference of renal function, as demonstrated usually by DTPA scan, is used to determine early renal damage before any deterioration of renal function. Blyth et al. [44] chose 35% relative function of the hydronephrotic kidney as the cnt-off point for early surgical intervention. Ransley et al. [39] chose 40% and Kass and Fink-Bennett [45] chose 45%. Note that the relative differences of function were chosen arbitrarily. Secondly, the decision to operate has traditionally been made when the T1/2s and washout pattern are in the so called "obstructive" range, noting that there is substantial evidence to suggest that these T1/2s and patterns on diuretic renogram can be misleading. Thirdly, Koff and Campbell [38] have proposed assessing all kidneys over a period of time and determining if there is worsening function or I-IN. This is the approach that we currently use to determine the timing of surgery if required. In addition, worsening morphology, accelerated contralateral normal kidney growth, failure to improve function, grossly obstructed renograms and positive Whitaker studies are parameters that have been suggested as guidelines to base one's decision to operate. Antibiotic prophylaxis from birth is recommended in preventing infection and possible deterioration. Initially,

508 amoxicillin is the antibiotic of choice in our department. Oral penicillin G is an alternative. After 4 - 6 weeks, antibiotic coverage is switched to trimethoprim and/or sulfamethoxazole. Prophylaxis has been proven efficacious and safe. Elucidating the natural history of H N as previously discussed has changed the trend of management. Conservative management of the hydronephrotic kidney is the most c o m m o n mode o f treatment of asymptomatic HN [46]. Conversely, in cases of solitary kidney with HN, early surgery is more commonly performed. Mayor et al. [47] and Perez et al. [48] have advocated surgery for H N that should not be delayed beyond the neonatal period. This was partly based on finding postoperative GFRs improving dramatically. This is now thought to only reflect expected maturation in renal function with age. The clinical support for operating to save potential renal function has been lacking.

Discussion HN is an anatomical description. Functionally, there exists a spectrum of obstruction from complete to partial to nonobstructive. HN therefore can be referring to any of these entities. Overall, in neonates and infants with asymptomatic HN, the natural history is becoming clearer. HN appears to be a relatively "benign" disease [37]. Conservative management has been shown not to lead to any permanent deterioration of kidney function [37]. Operative repair will become necessary in less than 10% of neonates. The proper diagnosis and follow-up of neonates and HN is a topic of hot debate. Presently, when an infant or child presents with asymptomatic HN, surgery, observation and a combination of the two is the management that is offered. With the elucidation of the natural history of HN, we now know that there is a significant number of children undergoing unnecessary operations. There is no present method of selecting the hydronephrotic kidney that would benefit from surgery. The few who need surgery benefit because of the 95% success rate of pyeloplasty and equally high success rate upon correction of UVJ obstruction. The controversy of conservative versus surgical management hinges on the assessment of renal function. The real issue is identifying patients at risk for renal damage. It is important to further research a better technique of assessing the hydronephrotic kidney than waiting for renal functional deterioration or arbitrarily choosing 10%, 20% or 30% relative renal functional difference as the threshold of renal damage that will trigger an operation. The kidney possesses a renal reserve that allows it to increase its filtrate rate in response to a variety of conditions [49]. The kidneys' response to obstruction is now thought to be primarily hemodynamic [8, 9] and involves multiple secondary messengers and, therefore, we are currently investigating pharmacologically "boosted" renograms with agents such as captopril and indomethacin. We are also performing "protein-load" renograms which reveal differences in maximal GFR using renal functional reserve as an indicator of renal obstructive vasoconstrictive dam-

age. We believe that studying these mechanisms will provide further insight into obstructive nephropathy. In spite of the tremendous controversy generated in HN, conservative management appears to be the most c o m m o n mode of treatment. When to approach HN surgically is a clinical dilemma that must be determined for the time being by the comfort zone of the individual urologist. Future research in determining the exact pathophysiology of partial obstruction, possibly preventing the sequelae of obstruction, and discovering new more sensitive diagnostic techniques capable of sensing early renal damage is urgently needed and is ongoing.

References 1. Homsy YL, Koff SA (1988) Problems in the diagnosis of obstruction in the neonate. In: King LR (ed) Urologic surgery in neonates and young infants. Sannders, Philadelphia, pp 77-94 2. Thomas DF (1990) Fetal uropathy. Br J Urol 66:225-231 3. Livera LN, Brookfield DS, Egglinton JA, Hawnaur JM (1989) Antenatal ultrasonography to detect fetal renal abnormalities: a prospective screening programme. BMJ 298: 1421-1423 4. Helin I, Persson PH (1986) Prenatal diagnosis of urinary tract abnormalities by ultrasound. Pediatrics 78:879-883 5. Grignon A, Filiatrault D, Homsy YL, Robitaille F, Filion R, Boutin H, Lebland R (1986) Urinary tract dilatation in utero: classification and clinical applications. Radiology 160:645-647 6. Homsy YL, Saad 17, Laberge I, Williot R Pison C (1990) Transitional hydronephrosis of the newborn and infant. J Urol 144: 579-582 7. Klahr S (1991) New insights into the consequences and mechanisms of renal impairment in obstructive nephropathy. Am J Kidney Dis 18:689-699 8. Klahr S, Purkerson ML (1994) The pathophysiology of obstructive nephropathy: the role of vasoactive compounds in the hemodynamic and structural abnormalities of the obstructed kidney. Am J Kidney Dis 23:219-223 9. Passerini-Glazel G, Koff SA (1993) Hydronephrosis in the newborn: an experimental overview Dialogues Pediatr Urot 16:1-8 10. Fichtner J, Boineau FG, Lewy JE, Sibley RK, Vari RC, Shortliffe LM (1994) Congenital unilateral hydronephrosis in a rat model: continuous renal pelvic and bladder pressures. J Urol 152: 652-657 11. Fitzpatrick JM, Ryan JPC, Geraghty JG, Lennon GM (1993). The effects of renal mad ureteric function on partial nreteric obstruction. Dialogues Pediatr Urol 16:4-6 12. Chevalier RL (1993) Renal hemodynamics and the renin angiotensin system in experimental ureteral obstruction. Dialogues Pediatr Urol 16:6-8 13. Bogaert GA, Mevorach RA, Kogan BA (1994) Renal hemodynamic and functional effects of 10 days partial urinary obstruction in the fetal lamb. J Urol 152:220-225 14. Lewy JE (1993) Congenital hydronephrosis in an animal model. Dialogues Pediatr Urol 16:3-4 15. Bowie JD, Rosenberg ER, Andreotti RE Fields SI (1983) The changing sonographic appearance of fetal kidneys during pregnancy. J Ultrasound Med 2:505-507 16. Chwalla R (1927) Ober die Entwicklung der Harnblase und der Primaren der Menschen mit besonderer Berticksichtigung der Art und Weise, in der sich die Uretern yon der Urnierengangen trennen, nebst Bemerkungen ~iber die Entwickiung der Mtillerschen Gange und des Mastdarms. Z Anat Entw Gesch 83:615 17. Ostling K (1942) The genesis of hydronephrosis particularly with regard to the changes at the ureteropelvic junction. Acta Chir Scand 86:1 18. Ruano-Gil D, Coca-Payeras A, Tejedo-Mateu A (1975) Obstruction and normal recanalization of the ureter in the human embryo.

509

19. 20. 21.

22.

23.

24.

25.

26.

27.

28. 29.

30.

31.

32. 33.

Its relation to congenital ureteric obstruction. Eur Urol l: 287-293 Allen TD (1970) Congenital ureteral strictures. J Urol 104: 196-204 Homsy YL, Williot R Danais S (1985) Transitional hydronephrosis: fact or fantasy? J Urol 136:339-341 Koff SA, Peller PA, Young DC, Pollifrone DL (1994) The assessment of obstruction in the newborn with unilateral hydronephrosis by measuring the size of the opposite kidney. J Urol 152: 596-599 Maizels M, Reisman ME, Flom LS, Nelson J, Fernbach S, Firlit CF, Conway JJ (1992) Grading nephroureteral dilatation detected in first year of fife: correlation with obstruction. J Urol 148: 609-614 Platt JF, Rubin JM, Ellis JH, DiPietro MA (1989) Duplex doppler US of the kidney: differentiation of obstructive from nonobstructive dilatation. Radiology 171:515 Keller MS, Garcia CJ, Korsvik H, Weiss RM, Rosenfield NS (1991) Resistive index ratios in the US differentiation of unilateral obstructive vs. nonobstructive hydronephrosis in children (abstract). Pediatr Radiol 21:462 Ordorica RL, Lindfors KK, Palmer JM (1993) Diuretic doppler sonography following successful repair of renal obstruction in children. J Urol 150:774-777 Mandell J, Peters DA, Estroff JA, Alfred EN, Benacekaff BR (1993) Human fetal compensatory renal growth. J Urol 150: 790-792 Hollowell JG, Altman HG, Snyder HM, Duckett JW (1989) Coexisting ureteropelvic junction obstruction and vesicoureteral reflux: diagnostic and therapeutic implications. J Urol 142:490 Weiss RM, Lytton B (1974) Vesicoureteral reflux and distal ureteral obstruction. J Urol 111:245 Krueger RE Ash JM, Silver MM, Kass EJ, Gilmour RF, Alton DJ, Gilday DL, Churchill BM (1980) Primary hydronephrosis: assessment of diuretic renography, pelvis perfusion pressure, operative findings, and renal and ureteral histology. Urol Cfin North Am 7 : 2 3 1 - 2 3 2 Eshima D, Taylor A (1992) Technetium-99m mercaptoacetyltriglycerine: update on the new 99mTc renal tubular function agent. Semin Nucl Med 2 2 : 6 1 - 7 3 Lambert R, Turpin S, Taillefer R, Leveille J (1993) An easy technique for GFR determination in children using 99mTc-DTPA, one plasma sample and external counting with a thyroid probe (abstract). J Nucl Med 34:5 Homsy YL, Mehta R Danals S (1988) Intermittent hydronephrosis: a diagnostic challenge. J Urol 140: 1222-1226 Kass EJ, Majd M (1985) Evaluation and management of upper urinary tract obstruction in infancy and childhood Urol. Clin. North Am. 12:133-141

34. Conway JJ (1992) Well tempered diuresis renography: its historical development, physiological and technical pitfalls and standardized technique protocol. Semin Nucl Med 2 2 : 7 4 - 8 4 35. Conway JJ (1992) The "well-tempered" diuretic renogram: a standard method to examine the asymptomatic neonate with hydronephrosis or hydroureteronephrosis. A report from combined meeting of The Society for Fetal Urology and members of The Pediatric Nuclear Medicine Council - The Society of Nuclear Medicine. J Nucl Med 33:2047-2051 36. Koff SA, Thrall JH, Keyes JW Jr (1979) Diuretic radionuclide urography: a non-invasive method for evaluation nephronreteral dilatation. J Urol 122:451-454 37. Koff SA, McDowell GC, Byard M (1988) Diuretic radionuclide assessment of obstruction in the infant: guidelines for successful interpretation. J Urol 140: 1167-1168 38. Koff SA, Campbell KD (1994) Nonoperative management of unilateral neonatal hydronephrosis: natural history of poorly functioning kidneys. J Urol 152:593-595 39. Ransley PG, Dhillon HK, Gordon I, Duffy PG, Dillon MJ, Barrat TM (1990) The postnatal management of hydronephrosis diagnosed by prenatal ultrasound. J Urol 144:584-587 40. Cartwright PL, Duckett JW, Keating MA, Snyder HM, Escala J, Blyth B, Heyman S (1992) Managing apparent ureteropelvic junction obstruction in the newborn. J Urol 148: 1224-1228 41. Flashner SC, Mesrobian HJ, Flatt JA, Wilkinson RH, King RL (1993) Nonobstrnctive dilatation of upper urinary tract may later convert to obstruction. Urology 42:569-573 42. Liu HY, Dhillon HK, Yeung CK, Diamond DA, Duggy PG, Ransley PG (1994) Clinical outcome and management of prenatally diagnosed primary megaureters. J Urol 152:614-617 43. Duckett JW (1993) When to operate on neonatal hydronephrosis. Urology 42:617-619 44. Blyth B, Snyder HM, Duckett JW (1993) Antenatal diagnosis and subsequent management of hydronephrosis. J Urol 149: 693-698 45. Kass EJ, Fink-Bennett D (1990) Contemporary techniques for the radioisotopic evaluation of the dilated urinary tract. Urol Clin North Am 17:273-289 46. Maizels M, Mitchell B, Kass E, Fernbach SK, Conway JJ (1994) Outcome of non-specific hydronephrosis in the infant: a report from the registry of The Society for Fetal Urology. J Urol 152: 2324-2327 47. Mayor G, Genton N, Torrado A, Guignard JP (1975) Renal function in obstructive nephropathy: long-term effect of reconstructive surgery. Pediatrics 56:740-747 48. Perez LM, Friedman RM, King LR (1991) The case for relief of ureteral pelvic junction obstruction in neonates and young children at time of diagnosis. Urology 38:195 49. Thomas DM, Coles GA, Williams JD (1994) What does the renal reserve mean? Kidney Int 45:411-416