Recoverability of renal function after relief of chronic partial upper urinary tract obstruction
Dr A.A. Shokeir Urology and Nephrology Centre, Mansoura University, Mansoura, Egypt.
Obstructive uropathy with resultant hydronephrosis is the eventual outcome of many urological disorders. Apart from accidental ligation or ureteric calculus, obstruction in humans is nearly always chronic and partial. The accurate prediction of the recoverability of kidney function after the release of chronic partial obstruction is of great clinical value to the urologist and nephrologist. If restoration or improvement of renal function appears probable, surgical relief of obstruction may be indicated, even though there has been a considerable initial loss of function. In patients whose kidney function is irreversibly damaged by the underlying obstructive process, it may be preferable to apply nonsurgical management or total removal of the disordered system by nephrectomy. The aim of the present paper is to discuss the different factors that may potentially affect recoverability and to review the currently available tests that have been proposed to predict recovery of renal function upon relief of chronic partial obstruction.
Factors that may affect recovery of renal function
Duration of obstruction
The degree of recovery of renal function after the release of unilateral ureteric obstruction correlates inversely with the duration of the obstruction [1,2]. In dogs there was rarely any return of renal function with release of unilateral complete ureteric obstruction after 40 days . However, the situation in humans is largely unknown. Function has been reported to return in clinical cases after more than 150 days of complete obstruction . The reason for the lack of correlation between the experimental data and the clinical situation is that the complete occlusion of the ureter by ligation obtained in animals is not comparable with the incomplete occlusions that are more likely to occur in clinical types of hydronephrosis.
Function of the contralateral kidney
In the past, theories of renal counterbalance and renal atrophy of disuse have considerably affected the treatment options of prolonged unilateral ureteric obstruction. It was commonly believed that once hypertrophy of the contralateral healthy kidney had fully developed the damaged kidney would not regain its function . Hydronephrotic atrophy was considered complete in 6 months, while compensatory hypertrophy was considered complete in 4–6 weeks . Animal as well as human studies challenged this theory and now it is well established that recovery of even a poorly functioning kidney is possible, despite compensatory hypertrophy of its partner. Nevertheless, the degree of recovery of the damaged kidney is influenced by the functional status of the contralateral kidney. Prominent recovery occurs under the stimulus of impaired function or removal of the opposite kidney [6–8].
Schirmer and Hendricks , studying metabolic aspects of unilateral hydronephrosis, reported that oxygen consumption in the obstructed kidney fell to 30% of that in the control after 2 weeks of complete ureteric occlusion. If the obstruction was then released, the oxygen consumption of the hydronephrotic tissue recovered to 70% of control values. However, if the contralateral ureter was subsequently obstructed, the recovery of the previously hydronephrotic kidney was further enhanced to 87% of the control.
Experimental studies in rats have indicated that age at the time of obstruction is an important factor determining the subsequent damage caused by ureteric obstruction. Provoost et al.  reported that immature kidneys appeared to be more vulnerable to the damaging effects of a complete ureteric obstruction than were adult ones. Furthermore, in a series of elegant experiments, Josephson et al. [11–13] showed that partial obstruction in the newborn rat induced more severe damage compared with weanling or pubescent rats. Finally, it should be noted that in rats with congenital hydronephrosis, the mere presence of hydronephrosis does not necessarily result in an impairment of GFR [14,15].
In children, many urologists feel that the potential for recovery of what seems to be a severely damaged renal unit is great . In a clinical study of 45 neonates with hydronephrosis, Koff and Campbell showed that even the most severely hydronephrotic functionally impaired kidneys had good potential for improvement . Bassiouny  reported remarkable recovery of renal function in 10 neonates with clinically palpable hydronephrotic nonvisualizing kidneys with a differential renographic clearance of <10%. In the light of previous comments, the unpredictable recoverability of obstructed hydronephrotic kidneys in young children must govern all action. However, there are studies to the contrary which show clearly that kidneys with <10% function may not perform in the long-term and a reconstructive procedure may be part of a two-stage nephrectomy .
It has been postulated that preservation of renal function in hydronephrosis is augmented through pyelo-lymphatic backflow. The reabsorption of renal pelvic urine into the lymphatics allows glomerular filtration to be replaced. It has been shown that if the lymphatics alone are ligated, there is no necrosis, only a parenchymatous degeneration of the tubular cells. Ligation of both the lymphatics and the ureter produces severe renal damage with necrosis, and destruction may be seen in several days instead of several months .
Compliance of the ureter and renal pelvis
If the ureter and renal pelvis are very compliant they would take the brunt of the back-pressure generated and protect the kidneys. However, if the ureter and pelvis are not compliant, particularly if the pelvis is intrarenal, there will be significant calyceal dilatation and atrophy of the renal cortex.
Other confounding factors
Return of function may well depend upon additional factors such as the presence or absence of infection, concurrent nephrotoxic medications, contrast materials and other nephrotoxic agents. Moreover, the component of dysplasia in the kidney plays a major influence in recovery of renal function, particularly in children. If an obstructed kidney has a significant dysplastic component it will not regain its function significantly after release of obstruction.
Methods of predicting recoverability
Excretory urography and grey-scale ultrasonography
Intravenous urography, although a primary diagnostic procedure, remains a qualitative examination and does not accurately indicate the functional mass of each kidney; IVU could not be used for predicting recovery in obstructive uropathy. Many cases with nonvisualizing kidneys on IVU regained reasonable function after the relief of obstruction in both children  and adults .
Grey-scale ultrasonography is of great help in the evaluating the degree of obstructive renal damage, through assessment of the thickness of the renal parenchyma. Moreover, grey-scale ultrasonography can indicate renal medical diseases through the echogenicity of the renal parenchyma. Provided that there is no renal medical disease, the greater the thickness of the renal parenchyma, the better the recovery of renal function after the relief of obstruction .
Sequential ultrasonograms in the paediatric age group may reveal a good potential for recovery if there is renal growth in the interim. It has been shown clearly that if the kidneys stop growing or do not grow as expected, the situation may be slightly worse.
Radioisotope renal scintigraphy has been suggested as a method to assess the potential reversibility in functional impairment of renal tissue damaged by obstructive disease. Several agents have been used for this purpose, e.g. 99mTc-DTPA) [18,20,22], iodine-orthoiodohippurate (131I-OIH) [4,23] and 99mTc-DMSA [24,25]. The results of various renal scans are conflicting because the different agents are handled differently by the kidney, and thus measure different aspects of renal function. 99mTc-DTPA is eliminated by glomerular filtration only and has been used as an index of GFR. 131I-OIH is not only filtered but also secreted in renal tubular cells and measures renal blood flow rather than GFR, although a fair correlation with GFR is usually present. After injection, 99mTc-DMSA accumulates in the cytoplasm of proximal tubular cells; 15–20% is eliminated in urine in 24 h and the tracer is used to assess functioning renal cortical tissue mass.
In early obstruction, renal tubular function is little changed and tubular agents such as 131I-OIH and 99mTc-DMSA may be useful in documenting decreasing renal function . However, in advanced stages of obstruction in which tubular function is greatly altered a glomerular agent, e.g. 99mTc-DTPA, is the choice for evaluating recoverable renal function. Sherman and Blaufox  showed that in patients with obstructive uropathy and nonvisualization of the kidney on hippuran scan, even if the kidney does not absorb hippuran, a potential for functional recovery exists. 99mTc-DTPA scans show residual function in advanced chronically obstructed kidneys, in which 131I-OIH, 99mTc-DMSA and IVU had all been negative.
Recently, a new technetium-labelled physiological analogue to 131I-OIH, 99mTc-MAG3, has been synthesized. Its biological properties are similar to those of hippuran, but its physical properties are superior. This agent combines the advantages of high renal extraction with an appropriate energy emission for the gamma camera. Both radiopharmaceuticals are filtered by the glomeruli and secreted by the renal tubules. Studies in normal, poorly functioning and transplanted kidneys have shown that 99mTc-MAG3 can replace hippuran in renography and no side-effects have hitherto been reported [27–29]. With these merits in mind, it becomes tempting to evaluate whether 99mTc-MAG3 scintigraphy can provide a sensitive tool for evaluating renal function with the intention of predicting potential functional recoverability in obstructive uropathy. The utility of this agent for this particular aspect has not yet been addressed.
Most investigators have used radionuclide quantification of renal function for assessing renal recoverability. Basically, two approaches have been taken to quantify renal function via renography. One has been to quantify function absolutely, i.e. to obtain a discrete value of GFR or effective renal plasma flow (ERPF). There is an almost general agreement among urologists that if the relative renographic GFR or ERPF has fallen to <10%, then very little recovery of the kidney function can be expected after de-obstruction, whereas when >10% kidney function can improve considerably [18,20,22–25]. The better the radionuclide uptake, the better the recuperation after de-obstruction. In a recent experimental study using 99mTc-MAG3 we showed that in dogs with a partial ureteric obstruction, resulting in an ERPF as low as 12% of the basal value, a good recovery is still possible upon removal of the obstruction .
Renography may be misleading in neonates and young children with severe hydronephrosis because blood flow is diminished, and the renal parenchyma is thin and displaced, resulting in inaccurate calculation . A poor renographic function alone does not justify nephrectomy in these patients. Therefore, the test should be used only to support surgical correction, but not to justify nephrectomy. Whenever there is a question of potential renal function in children, additional diagnostic studies are indicated.
Kalika et al.  used a qualitative approach for predicting recoverability of renal function, by comparing 131I-OIH renogram curves generated from the renal cortex and from the whole kidney. The results supported the hypothesis that when cortical curves appear more normal than total kidney curves there is a strong likelihood of postoperative improvement of renal function when the obstruction is relieved. Abnormal cortical curves are associated with a poor prognosis for renal functional improvement. Recoverability of kidney function in patients with acute obstruction was predicted in 89% and in those with chronic obstruction in 96%. The overall accuracy for acute and chronic obstruction was 94% . However, because of the occasional patient who may have improvement despite a negative prediction, Kalika et al. recommended that the test is better suited at present towards choosing patients in whom there is some possibility of surgical benefit .
Percutaneous nephrostomy (PCN)
A PCN performed before surgical correction can provide information about the potential for recovery of the affected kidney. Ransley et al.  studied newborns with PUJ obstruction and poor renal function by inserting a pigtail nephrostomy tube. The patients underwent a repeat 99mTc-DTPA scan after 3 weeks of drainage and proceeded to either nephrectomy (if function was <10%) or pyeloplasty (if function was >10%). A third of these patients showed useful functional recovery . Therefore, the authors recommended that nephrectomy should not be performed without a period of pigtail catheter drainage . Nevertheless, the same authors have recently changed their strategy; they no longer use nephrostomy drainage and proceed directly to nephrectomy if the 99mTc-MAG3 GFR is <10% .
Gillenwater  stated that the best method to determine the absolute degree of injury and recoverability is by temporarily relieving the obstruction with a nephrostomy tube and following renal function by creatinine clearance. However, the level of creatinine clearance above which obstruction should be corrected was not indicated. Cronan  advised allowing the renal unit sufficient time for recovery after placement of the PCN. Up to 8 weeks is necessary before the kidney has established its new baseline level of function. On the other hand, Bassiouny  did not recommend nephrostomy in neonates with poor renal function because it may be difficult to perform in neonates and eventually introduces infection, causes retraction of the renal pelvis and additional renal damage, and increases the risk of complications at the time of pyeloplasty, with more difficult surgery.
Bassiouny  recently reported experience in the management of 10 neonates who presented with nonvisualizing hydronephrotic kidneys caused by PUJ obstruction, with a differential 99mTc-DTPA uptake of <10%. The author recommended that the decision to remove or preserve a kidney in a neonate should be made at surgery after prior confirmation of the presence of obstruction. At operation, the parenchyma usually appears thin and stretched over a distended tense renal pelvis. The overall bulk of the stretched renal parenchyma should be assessed rather than its thickness, as decompression may be rewarded by the return of significant kidney function . Using the policy of immediate surgical repair after confirming the diagnosis, Bassiouny reported remarkable recovery of renal function (>100% on renal scintigraphy, albeit from low basal levels) in all the patients .
Huland et al.  proposed that hydronephrotic atrophy in adult rats after different grades of stable partial ureteric obstruction develops in two phases, a destructive phase followed by a steady-state phase. They showed that relief of obstruction in the destructive phase, but not in the steady-state phase, was able to ameliorate or prevent hydronephrotic atrophy. As the duration of ureteric obstruction in humans is usually unknown, the urinary enzymes were studied in rats after partial unilateral ureteric obstruction, to identify the destructive phase. The authors chose two markers of tubular damage, the lysosomal enzyme N-acetyl-glucosaminidase (NAG) and the brush-border enzyme gamma-glutamyl-transferase (gamma-GT). The NAG concentration, but less so that of gamma-GT, was higher in the urine of the obstructed kidney than in that of the contralateral control kidney, in the first 2 weeks after obstruction, and then returned to normal . These observations lead to the conclusion that the destructive phase after unilateral ureteric obstruction can be identified by the appearance of a high urinary tubular lysosomal enzyme content. The clinical implication is that the timing of relief of asymptomatic stable partial ureteric obstruction of unknown duration can be based on the concentration of urinary lysosomal enzymes.
Clinical studies in children  and adults  are promising, indicating that urinary enzymes may be helpful in the diagnosis of upper urinary tract obstruction, and in identifying the proper timing of its relief. However, urinary enzymes may be elevated in conditions other than urinary obstruction, among them being tubular damage due to nephrotoxic drugs, hypertension, diabetes and acute tubular necrosis . Moreover, high urinary enzyme levels were noted in patients with acute urinary tract infection . More work needs to be done to determine the clinical usefulness of urinary enzymes in predicting recoverability of renal function after release of obstruction.
Transforming growth factor-beta (TGF-β)
TGF-β is one of the growth factors elaborated by the kidney and which have been implicated in nephrogenesis. In the kidney, TGF-β affects mesenchymal differentiation and promotes extracellular matrix formation which results in collagen synthesis. Specifically in the urinary tract, TGF-β is elevated in the kidney after induced obstruction . A high level of TGF-β expression correlates with muscle hypertrophy and increased collagen deposition, which remodel the renal pelvis in response to obstruction. The lower level of TGF-β expression in chronic obstruction may reflect a limit of re-modelling, once a steady state has been achieved.
Seremetis and Maizels  recently showed that there is an increase of TGF-β mRNA expression in the renal pelvis after clinical and experimental PUJ obstruction. They have also shown that high levels of TGF-β mRNA expression correlated significantly with good clinical outcome. Other studies are needed to confirm these observations and to determine the correlation between the level of TGF-β and recoverability of renal function.
The resistive index (RI) obtained during Doppler ultrasonography (DU) has been recently described by Platt et al. . The noninvasive nature of the measurement of renal RI gives it considerable appeal in its potential application to patients with obstructive uropathy. We recently showed in an experimental study that recovery of renal function could not be predicted from the changes of RI before de-obstruction . However, we reported that a reversal of a previously elevated RI could be used as an early indicator that recovery of renal function is likely. This could be clinically applied by monitoring RI before and after the temporary release of obstruction via a PCN. In the future, we may be able to couple DU with the administration of vasoactive drugs to look for a reversal of a high RI to predict recovery after corrective surgery .
Other suggested methods
To our knowledge there are only three studies documenting the histological changes in children with PUJ obstruction [40–42]. In severe cases, PUJ obstruction may cause substantial renal damage. Steinhardt et al.  studied 20 patients undergoing nephrectomy for severe obstruction, of whom 15 had interstitial fibrosis with inflammation, 14 had glomerulosclerosis with inflammation, six had medullary dysplasia and three had glomerular cystic changes. Krueger et al.  previously showed that renal biopsy was normal in seven of 13 children undergoing pyeloplasty. In that series, two of four patients with a half-time drainage of <10 min had an abnormal biopsy, but four of six with a half-time drainage of >30 min had a normal renal biopsy. Elder et al.  retrospectively analysed the renal biopsies obtained during pyeloplasty in 55 children (mean age of 4.9 years, range 4 days to 19 years). Histological changes were compared with preoperative differential renal function; the changes were graded on a scale of 1–5. Differential function on diuretic renography correlated with histological grade, as kidneys with a normal histology had a higher differential renal function than abnormal kidneys. However, from those kidneys with a differential function of >40%, 21% had a histological grade 3 or 4 on renal biopsy. Conversely, from those kidneys with a differential function of <40%, a third were normal or had minimal histological changes. Thus, in about 25% of children with PUJ obstruction there is a disparity between preoperative differential renal function computed during diuretic renography and the findings on renal biopsy.
To the best of our knowledge, the correlation between the different degrees of histological changes during obstructive uropathy has never been correlated with the recoverability of renal function after relief of obstruction. Studies of such a relationship may help the urologist to predict the degree of recoverability of renal function after relief of obstruction, according to information gained via preoperative percutaneous renal biopsy.
Estimation of renal parenchymal volume
One of the arbitrary criteria used to indicate whether repair of obstructive uropathy is warranted is the measurement of parenchymal thickness by ultrasonography. It has been estimated that the presence of ≥1 cm of parenchyma is a good prognostic parameter . However, in hydronephrotic kidneys the parenchymal thickness is irregular and measuring it at one or more points does not reflect the true amount of remaining nephrons. Therefore, we propose that measuring the whole volume of the renal parenchyma by ultrasonography, CT or MRI may be beneficial in the context of predicting recoverability of renal function in obstructive uropathy.
Preoperative percutaneous renal biopsy and estimating the volume of renal mass provide both qualitative and quantitative information about the renal parenchyma, and may be a good combination in predicting recoverability of renal function.
Renal functional reserve
The concept of renal functional reserve was recently introduced to express the extent to which the kidney can respond to a metabolic or haemodynamic overload. The idea of the test is to measure the selective renal function in basal conditions by radioisotope renography; the kidney is then placed under stress by inducing diuresis, an injection of dopamine and an amino-acid load. The selective renal function is measured again under the stress; the renal functional reserve is determined by subtracting the renal function at baseline from that under stress. The details of the test were described previously . Some functional reserve can always be detected in this way, even in patients with significant advanced renal damage. Whether the estimated functional reserve correlates with the potential recoverability after de-obstructing a hydronephrotic kidney is difficult to ascertain at present.
Several important conclusions can be drawn:
(i) An IVU is unsuitable for predicting recovery of renal function in obstructive uropathy.
(ii) Most investigators agree that in the adult population, a kidney with a 99mTc-DTPA GFR of <10% of the total normal GFR (≈10 mL/min) is considered unsalvageable.
(iii) Renograms may be misleading in neonates and young children. Therefore, they should be used only to support the decision for surgical correction. The decision for nephrectomy should not be based solely on renograms, but other factors should also be considered.
(iv) It is difficult to differentiate unsalvageable from potentially salvageable kidneys when using a single test only. Under such circumstances, combining different tests may be helpful in reaching the appropriate decision.
(v) Extensive work needs to be done to establish the clinical usefulness of the new noninvasive predictors of recoverability of renal functional, e.g. urinary NAG excretion, renal TGF-β expression and the RI.
(vi) Correlations between recoverability of renal function and histological changes secondary to obstruction, e.g. renal mass as determined by ultrasonography, CT or MRI, and renal functional reserve need to be studied.