We're unable to sign you in at this time. Please try again in a few minutes.
We were able to sign you in, but your subscription(s) could not be found. Please try again in a few minutes.
There may be a problem with your account. Please contact the AMA Service Center to resolve this issue.
Contact the AMA Service Center:
Telephone: 1 (800) 262-2350 or 1 (312) 670-7827  *   Email: subscriptions@jamanetwork.com
Error Message ......
Review Article |

Progression of Chronic Renal Failure FREE

Henry T. Yu, MD
[+] Author Affiliations

From the William Jennings Bryan Dorn Veterans Affairs Medical Center and the Division of Nephrology, Department of Medicine, University of South Carolina School of Medicine, Columbia. The author has no relevant financial interest in this article.

Arch Intern Med. 2003;163(12):1417-1429. doi:10.1001/archinte.163.12.1417.
Text Size: A A A
Published online

Chronic renal failure is characterized by a persistently abnormal glomerular filtration rate. The rate of progression varies substantially. Several morphologic features are prominent: fibrosis, loss of native renal cells, and infiltration by monocytes and/or macrophages. Mediators of the process include abnormal glomerular hemodynamics, hypoxia, proteinuria, hypertension, and several vasoactive substances (ie, cytokines and growth factors). Several predisposing host factors may also contribute to the process. Treatments to delay progression are aimed at treating the primary disease and at strictly controlling the systemic blood pressure and proteinuria. The role of antihypertensive agents, statins, and use of other maneuvers such as protein restriction and novel approaches are also discussed herein.

Figures in this Article

Chronic renal failure is characterized by a persistently abnormal glomerular filtration rate (GFR). It represents an evolving process that is initiated by various causes, all with the common end result of persistent and usually progressive damage of varying severity to the kidneys. However, the rate of decline, often referred to as progression, can vary substantially. The present article will discuss processes that affect progression after the initial renal insult has occurred.

Chronic renal failure is a common problem. In the third National Health and Nutrition Examination Survey done from 1988 to 1994, 3% of the US adult population was found to have elevated serum creatinine values.1 Once the renal failure is well established, the rate of progression can be estimated, although limitations exist.29

Many features are common to progression of renal failure of various causes, and the final histologic appearance is one of glomerulosclerosis, interstitial fibrosis, and loss of native renal cells. Nevertheless, the causes of chronic renal failure are heterogeneous, and the mechanisms and locations of the initial injury may vary. Different animal models emphasize different aspects of the pathophysiologic characteristics and only incompletely replicate clinical disease.

Several morphologic features are prominent: fibrosis; loss of normal renal cells, mainly by apoptosis; and infiltration by monocytes and/or macrophages. These represent the end result of the constant interplay between vasoactive substances (ie, cytokines and growth factors).

Impairment of renal function correlates better with the extent of tubulointerstitial injury than with histologic glomerular injury.10 Interstitial fibrosis results from increased synthesis and decreased breakdown of extracellular matrix (ECM). The abnormal ECM contains an excess of normal components such as fibronectin, laminin, proteoglycans, and type IV collagen. Apart from the evident histologic changes, alterations in the ECM composition also change the ways the cells interact with the ECM, and these in turn affect gene regulation in response to specific growth factors. The details remain an active area of inquiry.11

Myofibroblasts (cells containing features of smooth muscle cells and of fibroblasts) are involved in the fibrogenic process and can secrete alpha 2(I) and alpha 2(III) collagens and fibronectin. Their origins vary because several types of intrarenal cells can transdifferentiate into myofibroblasts.12,13 Moreover, in animal models, injured tubular cells also contribute to the development of interstitial fibrosis.14

Physiologic cell death is a normal event in tissue homeostasis and is important for removal of unnecessary or damaged cells. In the context of renal disease, the balance between cell proliferation and apoptosis plays a critical role in maintaining an optimal number of cells after an insult.15,16 In chronic renal failure, there is a loss of normal resident cell population thought to be caused by a combination of abundant proapoptotic stimuli and diminished antiapoptotic stimuli. Examples of the former are transforming growth factor β (TGF-β), tumor necrosis factor (TNF), Fas ligand (FasL), and interferon α. At the same time, the normal ECM, probably through interaction with cell-surface β-integrin receptors, inhibits apoptosis. Because the normal ECM becomes replaced by an abnormal one, its antiapoptotic effect is lost.17,18 The end result is a decreased population of the normal glomerular and tubular epithelial cells.

The Fas apoptosis pathway is initiated by the binding of FasL to Fas, which triggers a cascade of intracellular signals that results in apoptotic deletion of Fas-bearing target cells.19 Fas and FasL are constitutively expressed in renal tubular cells.20,21 Animal studies suggest that up-regulation of Fas in the tubules in response to cytokines favors its binding to FasL located on adjacent cells and thus leads to apoptosis.21,22 Tissue hypoxia from decreased perfusion of the microvasculature in chronic renal failure also stimulates Fas-mediated apoptosis.23 In addition, podocyte apoptosis may play an early role in progression of diabetic nephropathy and in focal segmental glomerulosclerosis.2426

Monocytes and/or macrophages are recruited by cytokines, which are overexpressed in chronic renal failure. As a response to injury, overexpression of macrophage-colony stimulating factor by the tubules drives local macrophage proliferation in the kidney.27 Plasma levels of neopterin, a marker of monocyte activation, increase progressively with worsening clinical renal function.28 Macrophage infiltration in the interstitium correlates with the degree of renal dysfunction.2931 These cells amplify the response by producing more cytokines, which promote further fibrosis and apoptosis.16,27 Treatment modalities that decrease chemotaxis ameliorate renal failure.32 Experimentally, macrophage-derived cytokines, including interleukin (IL) 1β, IL-6, and TNF-α, inhibit expression of vascular endothelial growth factor (VEGF); this is probably partly responsible for impaired angiogenesis and capillary loss.33 Recently, mast cells in the interstitium have been found to correlate with severity of interstitial fibrosis in patients with various glomerulonephritides.34 That there is also association with its growth factor and myofibroblasts suggests that these cells may be involved in progression of interstitial fibrosis as well.34

For purposes of the present discussion, mediators are factors and processes that perpetuate renal dysfunction after an initial insult of sufficient severity has occurred. They usually occur as a consequence, no matter how remote, of the initial renal damage (Figure 1).

Place holder to copy figure label and caption

Pathophysiologic pathway of chronic renal failure.

Graphic Jump Location

In rats subjected to subtotal nephrectomy, compensatory hyperfiltration of the spared nephrons helps to maintain overall GFR. However, this adaptation also leads to glomerular hypertension, proteinuria, and progressive chronic renal failure.35 The stretching of the capillary tuft also stretches the adjacent mesangial cells, which induces mesangial cell proliferation and glomerulosclerosis at least partly by overexpression of cytokines such as platelet-derived growth factor (PDGF)36 and monocyte chemoattractant protein 1.37

Early diabetic nephropathy is well known to be associated with an elevated GFR.3840 Experimentally, increased glomerular capillary hydraulic pressure and hyperfiltration occur (despite normal systemic pressures) due to a proportionally greater reduction in the afferent relative to efferent arteriolar resistance.41 Indirect measurements suggest that glomerular capillary hypertension is present in human patients with diabetes as well, and there is also a correlation between urine albumin excretion and the glomerular pressure but not systemic pressure. Thus, as in the remnant kidney model, the glomerular hypertension in diabetic nephropathy itself propagates chronic GFR decline, at least partly by increasing protein leakage across the glomerular capillaries into the Bowman space.42 Apart from increased glomerular capillary hydraulic pressure, cytokines activated by injury may counteract tonic mesangial cell contraction and also contribute to hyperfiltration.43

Hyperfiltration has not been well studied in nondiabetic human renal disease, and the extent to which this occurs in humans is not known for certain. Keller et al44 recently found that white hypertensive patients have fewer but larger glomeruli, suggesting compensatory hyperfiltration. Modeling of human lupus nephritis also suggests that hyperfiltration occurs and is indeed beneficial, at least in the short term.9 Studies of patients who have undergone unilateral nephrectomy have shown no deterioration in renal function.4547 However, there may be a critical renal mass below which hyperfiltration becomes detrimental. Patients with greater than 50% loss of renal mass have been shown to have a long-term increased risk for proteinuria and renal insufficiency.48


Hypoxia has been regarded as a potential cause as well as effect of progression. There is loss of the postglomerular intertubular capillaries in chronic renal failure of various causes.49 Glomerular sclerosis is thought to contribute to this by decreasing downstream tubular blood flow. Expansion of the interstitial space may also diminish capillary perfusion of the tubules.50 The resultant hypoxia favors release of proinflammatory and profibrotic cytokines. Experimentally, this is associated with increased expression of the antiangiogenic factor thrombospondin 1 and decreased expression of the proangiogenic factor VEGF, which may impair angiogenesis and further propagate the hypoxia.33

Cellular hypoxia prevents degradation of the transcription factor hypoxia-inducible factor 1, which then becomes available to bind to hypoxia-response elements in genes that are switched on by hypoxia.51 One such hypoxia-response element has been found in the tissue inhibitor of metalloproteinase 1 promoter.52 Apart from this, hypoxia has been found to increase expression of endothelin (ET) 153 and collagen alpha 1(I) and to decrease expression of collagenase.52 Tubular hypoxia also favors expression of Fas in the tubular cell membrane and apoptosis.23


Proteinuria occurs as a result of glomerular capillary hypertension and damage to the permeability barrier in the glomerulus. Protein leaking across the glomerulus is taken up by the proximal tubule cells by endocytosis. This causes protein overload on the proximal tubular cells, leading to increased activation of the intrarenal angiotensin-converting enzyme (ACE)54 and also, either directly or via activation of transcription factors,55 to abnormal production of the following cytokines: ET-1, monocyte chemoattractant protein 1, and RANTES (regulated on activation, normal T-cell expressed and secreted).56 The cytokines favor fibrosis, apoptosis, and monocytic infiltration, further propagating the process.

In proteinuric animals, there is also direct translocation of growth factors such as TGF-β and hepatocyte growth factor directly from plasma into tubular fluid. These then interact with receptors located at the apical membrane of tubular cells to promote interstitial fibrosis.57,58

Specific protein metabolites may also be involved in the progression of chronic renal failure. Indoxyl sulfate, one such metabolite, has been shown to increase glomerulosclerosis in animals59 by increasing renal TGF-β synthesis, both directly and by promoting the expression of intercellular adhesion molecule 1, the latter leading to monocyte infiltration.60 Elevated urinary levels of this metabolite also correlate with a more rapid progression in humans.61 Animal and human studies have shown decreased serum and urinary indoxyl sulfate levels by using AST-120, an oral adsorbent.62,63

In the presence of impaired glomerular permeability, transferrin in association with iron also enters the tubular lumen and is taken up by the proximal tubule cells. Such accumulation has been demonstrated in human chronic renal disease.64 In vitro evidence using human proximal tubular epithelial cells indicates that iron-mediated lipid peroxidation65 and complement activation by the apotransferrin component66 contribute to toxic effects.

Consistent with its role in pathophysiology, proteinuria is a strong predictor of clinical progression of renal disease. The rapidity of GFR decline is proportional to the severity of proteinuria.67

Systemic Hypertension

Systemic hypertension is a frequent accompaniment to chronic renal disease. Sodium, volume excess, and activation of the renin-angiotensin-aldosterone system in patients with chronic renal failure all cause hypertension. In addition, afferent stimuli from the kidneys may activate the sympathetic nervous system and contribute to the elevated pressure.68 Hypertension itself accelerates decline in renal function,6971 likely due to the associated increased glomerular capillary hypertension.

Complement Activation

There is at least 1 animal model in which proteinuria is ameliorated by inhibition of complements.72 Clinically, patients with proteinuria have been shown to excrete complement degradation products into the urine.73 Because of abnormal glomerular permeability, complement can enter the tubular lumen and initiate formation of the C5b-9 membrane attack complex.74 Exposure of tubular cells to an unidentified component of serum protein have also been shown to cause increased synthesis and release of complements by the cells themselves, predominantly toward the basolateral component.75 It has also been suggested that hyperammoniagenesis resulting from intratubular catabolism of excessive protein load also leads to complement activation and consequent interstitial scarring.75

Angiotensin II

Angiotensin II (AII) is formed by progressive cleavage of angiotensinogen. The kidneys contain all the machinery necessary to generate AII locally.76 This local intrarenal renin-angiotensin system is regulated independently of the systemic one and plays a critical role in renal autoregulation and pathophysiologic developments.77 Enhanced sensitivity to effects of locally produced AII is present in diabetic rats78 and has been postulated in an animal model of nondiabetic renal failure.79

Apart from its hemodynamic effects, AII also stimulates expression of fibronectin80 and several other downstream cytokines and growth factors that favor fibrogenesis and recruitment of macrophages. Examples are TGF-β, plasminogen activator inhibitor 1 (PAI-1), aldosterone, ET,81 osteopontin,82 and possibly the transcription factor nuclear factor κB.83

Other Chemical Mediators

As noted above, a growing list of vasoactive substances (ie, cytokines and growth factors) have been shown to be involved in progression of renal disease8495 (Table 1). Cytokines and growth factors gain access to the kidneys by multiple pathways. They can be synthesized elsewhere and be ultrafiltered across the glomeruli and act on tubular cells through apical receptors.57 They can also be secreted by resident renal cells in response to various stimuli and by infiltrating monocytes. Moreover, chemical mediators probably do not act in isolation; they often mediate the expression or release, or modulate the effect, of other mediators.9698 There is overlap,57 and the net effect can can be either beneficial or detrimental, depending on the physiologic context. There is also no universal agreement on the overexpression and underexpression of all cytokines among different models of renal disease studied by various investigators.

Table Graphic Jump LocationTable 1. Selected Cytokines and Growth Factors Involved in Progression of Chronic Renal Failure

Transforming Growth Factor β. Transforming growth factor β is the most well-studied and probably the most significant of the fibrogenic cytokines. It is produced by resident renal cells and by infiltrating monocytes. Production is stimulated by various chemical stimuli such as AII, elevated plasma glucose, and IL-1,13 as well as by mechanical stretching of mesangial and renal tubular cells.99 Directly and by increasing expression of other cytokines, TGF-β1 favors deposition of new ECM and decreases its degradation. It also favors monocytic/macrophage infiltration, transdifferentiation of tubular cells into myofibroblasts,100 and podocyte apoptosis.25

Among the downstream cytokines affected by TGF-β1 are PAI-1, hepatocyte growth factor, connective tissue growth factor, and nitric oxide. The causal role of this cytokine was evidenced by a recent study using an animal model of type 2 diabetes mellitus, in which chronic administration of a monoclonal antibody against TGF-β prevented the development of renal insufficiency, expression of matrix components, and histologic changes.101

Plasminogen Activator Inhibitor 1. There are 2 major systems involved in degradation of ECM: the plasminogen system and the matrix metalloproteinases. Both systems are activated by tissue-type plasminogen activator. Not detectable in the normal kidney, PAI-1 is expressed in the chronically damaged kidney and inhibits the ability of tissue-type plasminogen activator to convert plasminogen to plasmin and tissue matrix metalloproteinases to the active form. In addition, PAI-1 may also inhibit the endothelial isoform of nitric oxide synthase and thereby locally decrease nitric oxide production.

Expression of PAI-1 is increased by AII, angiotensin IV,102,103 TGF-β,104 and aldosterone.105 Serum samples from patients with azotemia have been shown to favor cytokine-induced secretion of PAI-1.106 It has also been hypothesized that there may normally be a tonic inhibition of PAI-1 transcription involving cell-matrix interaction. Loss of this inhibition, mediated either by cytokines or by loss or change of the normal matrix,107 favors ECM deposition.

Nitric Oxide. Nitric oxide is formed by the oxidation of L-arginine into L-citrulline. Three isoforms of the nitric oxide synthase catalyze this reaction: constitutive endothelial nitric oxide synthase, neuronal nitric oxide synthase, and inducible nitric oxide synthase. Nitric oxide inhibits mesangial cell proliferation and ECM synthesis and may limit capillary permeability. In rat models, nitric oxide inhibition results in proteinuria,79,108 increased blood pressure, and decreased GFR independent of renal AII levels.79 Collagen I expression is also increased, independent of systemic hemodynamics.81 It has been suggested that nitric oxide inhibits collagen I expression in the renal vasculature, and the detrimental effects of endogenous AII are increased in states of low nitric oxide availability. A polymorphism of the endothelial nitric oxide synthase gene has been described and found to be associated with development of diabetic nephropathy, suggesting that the associated lower mean plasma nitric oxide level may translate into decreased suppression of ECM synthesis.109

Total nitric oxide production has been shown to be low in chronic renal failure in most, but not all, studies. The major source of endogenous arginine is normally the proximal tubules. Whole-body L-arginine synthesis of arginine remains normal in hemodialysis patients110 and may reflect compensatory extrarenal synthesis. The extent to which local arginine availability may affect nitric oxide production locally is unclear. Parathyroid hormone, which is elevated in chronic renal failure, down-regulates nitric oxide synthase expression.111 Various substances overexpressed in renal failure (PAI-1,112 asymmetric dimethylarginine,113 PDGF,114 TGF-β,115 and ET-1116) can inhibit nitric oxide synthase, thereby diminishing conversion of L-arginine to nitric oxide. In patients with diabetes, scavenging of formed nitric oxide by advanced glycosylation end products may also contribute to the decreased nitric oxide levels.

On the other hand, cytokines activated by injury may inappropriately activate inducible nitric oxide synthase.117 It has been proposed that the resultant relaxation of tonic mesangial cell contraction leads to hyperfiltration and contributes to progression of renal failure.43

Aldosterone. Aldosterone levels are often elevated in normokalemic patients with chronic renal failure.118,119 Animal studies suggest that aldosterone may mediate progression of chronic renal failure.120 Subtotal nephrectomy in rats results in adrenal hypertrophy and elevation of plasma aldosterone level. Arterial hypertension, proteinuria, and glomerulosclerosis develop; these are ameliorated by the combination of an angiotensin receptor blocker (ARB) and an ACE inhibitor. Infusion of exogenous aldosterone restores the deleterious effects of subtotal nephrectomy, despite the concomitant administration of the ARB and the ACE inhibitor.120 These can be prevented by adrenalectomy.121

Aldosterone has also been shown to up-regulate ACE messenger RNA expression in cultured neonatal rat cardiocytes, thus completing a positive feedback mechanism.122 Aldosterone increases PAI-1 expression and may induce renal injury through this mechanism.105,123 Alternatively, there is experimental evidence that aldosterone may mediate renal vascular damage independent of its effects on blood pressure.124,125 It has been hypothesized that fibrosis may then be a secondary effect of the vascular damage.125

Endothelin. The ET system consists of 2 receptors, 3 ligands, and 2 activating peptidases. The 2 mammalian receptors are labeled ETA and ETB. The 3 ligands are ET-1, ET-2, and ET-3. All are constitutively synthesized and released by glomerular and tubular cells. Endothelin expression is favored by AII,126 IL-1, TGF-β, glucose, and hypoxia,53 among others.

Several actions mediated by ET may play a role in the progression of renal failure: blockage of inducible nitric oxide synthase transcription via the ETA receptor; increased expression of the collagen 1 gene81; vascular remodeling127; mediation of proteinuria128; macrophage chemotaxis; and stimulation of interstitial fibroblast proliferation and ECM synthesis. Endothelin may also mediate renal activation of nuclear factor κB, which in turn regulates the transcription of other genes involved in renal injury.83 Experimentally, the use of an ET receptor antagonist in combination with an ACE inhibitor improves proteinuria and histologic changes beyond what occurs with either agent alone. This suggests that ETs play a role in at least some animal models.83,129 In the rat subnephrectomy model, blockade of the ETA receptor reduces proteinuria more than does nonselective blockade.130 It has also been shown that the ETA receptor, but not the ETB receptor, mediates salt sensitivity of AII-induced hypertension in the rat.126,131

Platelet-Derived Growth Factor BB. Platelet-derived growth factor has been implicated in the progression of renal injury. It stimulates mesangial proliferation and increases ECM synthesis. Furthermore, overexpression of this cytokine, as well as its receptor, in the glomeruli and tubular and interstitial compartments has been demonstrated experimentally.94,132 Levels of PDGF are increased in response to AII, lipoproteins, ET, and other cytokines.

Two chains, PDGF-A and PDGF-B, form the active homodimers or heterodimers. Experimentally, PDGF-BB, but not PDGF-AA, induces renal myofibroblast transdifferentiation and tubulointerstitial fibrosis.133 Platelet-derived growth factor BB has also been shown to increase the expression of type III collagen by tubular cells and by myofibroblasts.57,134 In rat anti-Thy1.1 nephritis (a model of mesangioproliferative glomerulonephritis), transient antagonism of PDGF after disease induction prevented development of glomerulosclerosis, tubulointerstitial damage, and collagen accumulation.135

Numerous host factors have been associated with progression of renal failure. The most important ones are discussed below.

ACE Gene Polymorphism

A 287–base-pair fragment in intron 16 in the ACE gene can either be present (the I allele) or absent (the D allele). Presence of the D allele is associated with elevated systemic ACE levels while the I allele is associated with the opposite effect. Both alleles are codominant such that the 3 resulting genotypes (II, ID, and DD) are associated with low, intermediate, and high amounts of circulating ACE, respectively. This relationship holds true for intrarenal ACE as well.136

The D allele has been found in numerous studies to be either indifferent or deleterious to the progression of renal failure, as noted in a recent review.137 A study in patients with IgA nephropathy indicated that the risk associated with the D allele was most apparent in patients without proteinuria or hypertension, suggesting a weaker effect compared with these 2 risk factors.138

There remains controversy as to whether the presence of the D allele affects response to antiproteinuric therapy with ACE inhibitors. Conclusions from retrospective studies on nondiabetic nephropathy are divided, with some finding a poorer response with the I allele, and others with the DD genotype.137 A prospective study on type 1 diabetes mellitus has shown better response of albuminuria to ACE inhibitor treatment among those with the II genotype.139

Other polymorphisms have also been studied, albeit to a lesser extent than that involving the ACE gene. Examples include the angiotensinogen M235T polymorphism, the chymase gene CM A/B polymorphism, and the angiotensin receptor A1166C polymorphism.140 Of these, the M235T polymorphism is thought to contribute to risk of developing chronic renal failure, though this remains undefined.140


Smoking has been found to be associated with progression in diabetic nephropathy,141,142 primary renal disease,143 and severe hypertension.144 Smoking is a risk factor for proteinuria independent of the presence of diabetes and blood pressure145 and may contribute to progression because of the associated proteinuria. Elevation of ET-1 levels146 and acceleration of atherosclerosis and ischemic nephropathy147 may also be contributory. There are no prospective studies addressing whether smoking cessation ameliorates progression of renal failure.

African American Descent

In a prospective cohort study of previously untreated nondiabetic men with hypertension enrolled in the Multiple Risk Factor Intervention Trial (MRFIT), effective blood pressure control was associated with stable or improving renal function in nonblacks but not in blacks.148 A subsequent article149 involving all participants of the above MRFIT study found an increased risk of end-stage renal disease among blacks independent of several other factors. Analysis of the baseline characteristics of the Modification of Diet in Renal Disease study also identified black race as an independent predictor of a faster GFR decline.150

Socioenvironmental factors and genetic background have been proposed to account for the tendency toward excessive disease progression in African Americans. Part of the greater susceptibility may be from increased cytokine activation: ET-1151 and TGF-β152 levels have been found to be more elevated in African Americans with hypertension than in their white counterparts.

Diabetes Mellitus

Men with diabetes have been found to have a higher incidence of end-stage renal disease ascribed to nondiabetic causes, even after accounting for age, ethnicity, income, blood pressure, cholesterol, and history of coronary artery disease.153

Male Sex

Various studies have suggested that nondiabetic renal diseases progress more rapidly in men.154158 However, the studies vary widely in design and methodology. Not all studies have clearly shown the effect of sex to be independent of other factors such as proteinuria, severity of hypertension, and smoking history. Various mechanisms have been reviewed elsewhere159,160 and include increased response to AII in men161 and estradiol's ability to reverse TGF-β1–mediated fibrogenesis.162


Chronic renal failure is associated with elevation of triglyceride levels, oxidized low-density lipoprotein, lipoprotein (a), and decreased apolipoprotein (a). Renal failure itself may also promote hyperlipidemia by down-regulating the expression of the enzyme lecithin:cholesterol acyltransferase in the liver and its activity in the plasma.163

Experimentally, hypercholesterolemia and hypertriglyceridemia can each promote proteinuria and tubulointerstitial injury,164 while treatment aimed at decreasing lipid levels ameliorates the rate of progression.165 Putative mechanisms of damage include stimulation of reactive oxygen species, inhibition of nitric oxide, modulation of mesangial growth and proliferation, monocyte infiltration,32 and stimulation of growth factor and cytokine release.166168

In humans, various lipid abnormalities have been associated with the development of new renal insufficiency169 and progression of established renal disease.150,170 However, a definite causal relationship is equivocal. Moreover, the component(s) of the dyslipidemic milieu most responsible for progression is not clearly defined.169,171 Elevation of total cholesterol levels, low high-density lipoprotein, elevated triglyceride levels, and apolipoprotein B–containing lipoproteins have all been implicated.169,172,173

Recreational Drug Use

The use of heroin and other opiates174 and of cocaine has been found to be associated with increased risk for development of end-stage renal disease.175 Cocaine use may exacerbate hypertensive nephrosclerosis through progression of renal ischemia. It is unclear whether heroin and opiate use is causally related to the increased risk or represent only a surrogate marker.

Prenatal Factors

Animal studies have shown a decrease in glomerular number with induced intrauterine malnutrition176 but not with spontaneous low birth weight.177 In humans, the number of glomeruli correlates directly with the birth weight.178 There is also a direct correlation between low birth weight and chronic renal disease, which appears to hold true across races.179181 While a resultant lowered renal reserve and any possibly compensatory glomerular capillary hypertension might theoretically accelerate progression to end-stage renal disease, it is not yet clear if the low birth weight itself is directly responsible for the increased incidence of chronic renal failure because hypertension182,183 and diabetes183,184 are also associated with retardation of intrauterine growth.185

The major steps involved in slowing the progression of renal failure are outlined in Table 2. More detail is provided below.

Table Graphic Jump LocationTable 2. Maneuvers to Slow the Progression of Chronic Renal Disease
Treatment of Underlying Disease

Treatment of some causes retards progression to chronic renal failure. Experimentally, antagonism of PDGF during the acute phase of an animal model of mesangioproliferative nephritis prevented functional and morphologic changes of chronic renal failure from developing.135 Clinically, this may involve treatment of acute disease, although this is not invariable. For example, even in established chronic diabetic nephropathy, euglycemia of 10 years' duration following pancreatic transplantation has been shown to reverse histologic renal lesions.186

At the other extreme, there is encouraging if early evidence that at least in some animal models, more specific downstream therapy may be useful for well-established chronic disease. Examples include use of relaxin (which decreases macrophage infiltration and interstitial fibrosis independent of hemodynamic effects187), VEGF,188 and aldosterone antagonists.189

Treatment of Hypertension

The renal benefit of treatment depends to a significant extent on the underlying proteinuria. This was borne out in the Modification of Diet in Renal Disease study, which consisted of 2 randomized clinical trials. Study 1 evaluated patients with GFRs of 25 to 55 mL/min, while study 2 evaluated patients with GFRs of 13 to 24 mL/min. In both studies, patients were randomized to groups with mean arterial pressure (MAP) goals of either 92 mm Hg or 107 mm Hg. The decline in GFR was found to be slower in the more aggressively treated group overall, and the benefit was in direct proportion to the severity of the baseline proteinuria.190

In renal disease from type 1 diabetes mellitus, tighter blood pressure control, independent of use of ACE inhibitors, decreases proteinuria.191 A recent study from the Steno Diabetes Center showed that tight blood pressure control (MAP goal of 93 mm Hg) in this population can decrease the GFR decline to that found with normal aging.192

The UK Prospective Diabetes Study (UKPDS) and the Hypertension Optimal Treatment (HOT) study suggest that blood pressure can be aggressively yet safely lowered. The UKPDS showed a linear relationship between blood pressure and microvascular disease in patients with type 2 diabetes mellitus. This held true for average systolic blood pressures at least as low as 114 mm Hg.193 The HOT study showed that treating the blood pressure to as low as 120/70 mm Hg was not associated with any increase in cardiovascular events or mortality. Moreover, in the subset of patients with diabetes, cardiovascular events and mortality were lowest in the group assigned to a target diastolic blood pressure of 80 mm Hg or lower.194

Thus, to slow down GFR decline, the goal for patients without diabetes with significant proteinuria involving at least 1 g/d and for patients with diabetes is to lower the MAP to the low 90s. Based on the Modification of Diet in Renal Disease study data, Peterson and coworkers195 have suggested a MAP goal of 98 mm Hg or lower for patients with proteinuria involving between 0.25 and 1 g/d.

Treatment of Proteinuria

ACE Inhibitors. As many of the pathophysiologic changes associated with chronic renal disease are driven by AII, ACE inhibitors have become the logical and accepted choice for treatment. ACE inhibitors block the rate-limiting step in the formation of AII. Benefits have been shown in experimental and clinical settings. These drugs preferentially dilate the efferent arteriole, thereby hemodynamically decreasing glomerular hypertension and proteinuria.196,197 They also decrease proteinuria by preserving the integrity of component proteins of the slit diaphragm198,199 and by ameliorating podocyte foot process broadening.200

Numerous other salutary effects have been found. Experimentally, ACE inhibitors ameliorate monocyte/macrophage infiltration, TGF-β expression, fibroblast proliferation, differentiation into myofibroblasts,201 and development of interstitial fibrosis.202 Inhibition of at least some of these AII-mediated cytokine releases is probably due to decreased TGF-β1, which has been shown in patients treated with ACE inhibitors.203,204

Clinical benefits have been found for diabetic and nondiabetic renal disease.67,205207 In patients with diabetes, ACE inhibitors prevent progression of microalbuminuria, even in patients with controlled blood pressure. In a European study on type 1 diabetes mellitus, mean baseline systolic and diastolic blood pressure in the group randomized to receive lisinopril were 122 and 79 mm Hg, respectively. Adjustment for effects of systolic and diastolic blood pressure reduced, but did not eliminate, this benefit.206 A meta-analysis of 12 studies involving type 1 diabetes mellitus also came to the same conclusion.208

Even in the presence of chronic renal insufficiency, ACE inhibitors can be used.207,209,210 The benefits are proportional to the extent of proteinuria.67,210 Improvement in proteinuria is evident within the first 2 months.67,211 The preservation of GFR is directly proportional to the extent of lowering of proteinuria, which thus serves as a useful prognostic indicator.67,212

Angiotensin Receptor Blockers (ARBs). Because ACE can be formed by non–ACE-dependent pathways and because of intolerance to ACE inhibitors in many patients, ARBs have been increasingly used to delay progression of renal damage. Experimentally, ARBs have been shown to block fibroblast proliferation and synthesis of TGF-β.201 As in the case of ACE inhibitors, multiple mechanisms are responsible for the antiproteinuric effect. Angiotensin receptor blockers block expressions of cytokines like VEGF.213 They also have been found to normalize the glomerular nephrin deficiency214 and podocyte foot process broadening in diabetic animals.200,214 Animal studies suggest that the benefits are similar to those of ACE inhibitors.198,214

Recent clinical trials have shown that these agents diminish proteinuria215217 and protect against renal function decline among patients with type 2 diabetes mellitus and nephropathy.215,216 The 2 largest studies have been done on patients with relatively mild azotemia. However, results from studies using ACE inhibitors suggest that ARBs may be beneficial even in patients with more advanced chronic renal failure.

No large-scale comparative studies have been published comparing ARBs and ACE inhibitors regarding progression of chronic renal failure. However, small studies suggest comparable benefits in antiproteinuric effects,218,219 and no differences in rate of progression were found in a study of diabetes mellitus at 1 year.220 Animal data indicate that the combination results in better reduction of renal AII levels than either agent alone.221 Clinical studies have been conflicting as to whether a combination of both agents has additive antiproteinuric effects.218,222,223 However, the addition of ARBs to maximal ACE inhibition may reduce renal TGF-β1 production despite the lack of salutary effects on proteinuria.223,224 Unfortunately, no large prospective studies on rate of progression using combination therapy are currently available.

Calcium-Channel Blockers. Dihydropyridines do not have antiproteinuric effects.210,225 Nondihydropyridines, in contrast, appear to have some role, at least in diabetic nephropathy.226,227 There are no good data on its usage in nondiabetic nephropathies for retarding progression of renal disease, although plans for a multicenter study were recently published to investigate combined treatment with an ACE inhibitor and either a dihydropyridine or a nondihydropyridine calcium channel blocker, using proteinuria as an end point.228

β-Blockers. Though smaller studies had suggested that β-blockers may worsen the decline of GFR in patients with diabetes,227,229 the larger UKPDS-39 study found atenolol and captopril to be equivalent in limiting progression of albuminuria and azotemia over 4 years.230 Interestingly, a recent animal study found that blocking the sympathetic nervous system may be beneficial independent of antihypertensive effects.231 In a multicenter prospective study of nondiabetic African Americans with hypertensive renal disease, metoprolol was comparable with ramipril in reduction of proteinuria and in progression to end-stage renal disease.232 However, a composite end point and the overall rate of GFR decline was worse with metoprolol, suggesting that ACE inhibitors should still be the preferred agents in this population.

Protein Restriction

The role of protein restriction remains controversial. The largest prospective study to evaluate the role of protein restriction in retarding progression of renal disease failed to show a significant benefit.190 Subsequent secondary analysis of the data revealed a correlation between decreased protein intake and slower progression.233

Two relatively recent meta-analyses of low-protein studies have been performed.234,235 In both, nondiabetic patients with chronic renal failure were shown to benefit from protein restriction. However, it appears that the difference in rate of decline of GFR is small. Pooling the results of 13 randomized control trials, Kasiske et al235 estimated a difference of only 0.53 mL/min per year among those assigned to protein restriction, which may not be clinically meaningful. For diabetic chronic renal disease, the benefits seemed greater, but the pooled number of patients was small, totaling only slightly over 100 in both meta-analyses.

A caveat has to be noted. It has been shown that a low-protein diet decreases the filtered urine creatinine as well as urine creatinine secretion and affects the latter more than the former. Such changes can occur independent of changes in the GFR.8

In the clinical trial setting, at least, low-protein diets have been safe and have not been associated with hypoalbuminemia or other evidence of worsening malnutrition.190,236,237 Whether this is replicable outside of a research environment remains to be seen. It also remains unclear whether there is benefit in combining a low-protein diet with use of ACE inhibitors.

For patients with advanced renal failure (GFR <25) who are not undergoing dialysis, the K/DOQI guidelines238 currently recommend protein restriction to 0.6 g/kg per day, but allowing for a maximum of 0.75 g/kg per day. While this avoids generation of nitrogenous metabolites and ameliorates uremic manifestations, an outright benefit in delaying progression remains unproven.

Treatment of Hyperlipidemia/Use of Statins

Most of the studies have used statins directed at lowering cholesterol,239 although some work has also been published on the use of triglyceride-lowering agents.240 Apart from their lipid-lowering effects, statins also have other actions that may ameliorate the progression of chronic renal failure. Statins down-regulate TGF-β expression,241 interfere with intracellular signaling pathways,241 and prevent the activation of nuclear factor κB and substances downstream of TGF-β such as mitogen-activated protein kinases and connective tissue growth factor. Statins also possess antioxidant activity242 and some antihypertensive effect.243 A recent meta-analysis of trials that were mostly based on statins showed a modest benefit in slowing down GFR decline.239

Aldosterone Antagonism

In vitro, in the presence of AII, aldosterone further increases PAI-1 expression.105 This raises the possibility that aldosterone antagonism may have additional benefits beyond those of AII antagonism. Moreover, aldosterone suppression by ACE inhibitors alone may not be sustained.244 In different models of renal failure, however, aldosterone antagonism only partially reverses some, but not all, of the deleterious effects.120,125,245 A recently reported case series of 8 patients showed improvement in proteinuria by adding spironolactone to enalapril, though it is unclear if this occured independently of blood pressure changes.246


Erythropoietin receptors are present in the human kidney.247 It has thus been speculated that erythropoietin can exert cytokine effects on the kidneys and regulate their survival and proliferation.247 Apart from amelioration of hypoxia, this may have a salutary effect in progression of renal failure. Prospective clinical studies have been contradictory as to whether renal function is better preserved with erythropoietin.248,249

Binding Protein Metabolites

Indoxyl sulfate is a protein metabolite that promotes the progression of glomerulosclerosis in animal studies. It was recently shown in a small study that AST-120, an oral adsorbent that binds indole (the precursor for indoxyl sulfate in the gut), decreases serum and urinary levels of indoxyl sulfate and also improves the rate of decline of GFR in patients with chronic renal failure.62

Early Nephrology Referral

Studies on the effect of timing of nephrology referral on mortality have yielded conflicting results. However, late referral is associated with an increase in early morbidity.250 Specifically, early nephrology referral is associated with better predialysis care251 and more appropriate choice of angioaccess for eventual hemodialysis.251,252 It has also been found to be cost-effective.253

Other Modalities

Experimental Pharmacologic Agents. Endogenous atrial natriuretic peptide and brain natriuretic peptide cause vasodilation and natriuresis. The peptides are broken down by neutral endopeptidases. Since the active sites of neutral endopeptidases and ACE are structurally similar, vasopeptidases were developed to inhibit both enzymes.254 In animal studies, the vasopeptidase inhibitor omapatrilat was recently shown to be at least as efficacious as ACE inhibitors in decreasing proteinuria and glomerulosclerosis.255,256

Peroxisome proliferator–activated receptor γ agonists are currently used to enhance insulin sensitivity in type 2 diabetes mellitus. Other effects apparently independent of plasma glucose have been reported. Salutary effects include amelioration of histologic changes in an animal model of this disease257 and in nondiabetic glomerulosclerosis.258 A short study has also shown improvement of microalbuminuria in diabetic patients.259

Experimentally, variable but often beneficial effects have been reported by blocking harmful cytokines and vasoactive substances such as PDGF,135 ET,83,260,261 TGF-β,101 and epidermal growth factor.262 Other studies have used cytokines such as relaxin,187 hepatocyte growth factor,263,264 and bone morphogenetic protein 7.202 Antifibrotic agents265267 have also produced encouraging results.

Dietary Modification. Soy protein and flaxseed have shown benefit in some animal models.268271 Human studies are equivocal and are limited by small sample sizes and/or short follow-up. The longest prospective study to date, a crossover trial involving 6 months each of soy protein or animal protein, failed to show any change in GFR.272 This topic has been recently reviewed.273

Chronic renal failure is a common problem affecting a large number of people in the US population. Hemodynamic factors and various chemical mediators contribute to progression of chronic renal failure. Aggressive control of blood pressure and proteinuria, preferably with a regimen containing ACE inhibitors, remains the cornerstone of therapy. The role of protein restriction remains poorly defined but it does not appear to be generally useful. Statins may be useful. Novel agents and other modalities are at varying stages of development.

Corresponding author and reprints: Henry T. Yu, MD, Primary Care and Subspecialty Medicine (111), William Jennings Bryan Dorn Veterans Affairs Medical Center, 6439 Garners Ferry Rd, Columbia, SC 29209 (e-mail: henry.yu@med.va.gov).

Accepted for publication August 30, 2002.

I would like to thank Mattie M. Ashford, RN, for her help with this article.

Coresh  JWei  LMcQuillan  G  et al.  Prevalence of high blood pressure and elevated serum creatinine level in the United States. Arch Intern Med. 2001;1611207- 1216
Link to Article
Mitch  WEWalser  MBuffington  GALemann  J  Jr A simple method of estimating progression of chronic renal failure. Lancet. 1976;21326- 1328
Link to Article
Rutherford  WEBlondin  JMiller  JPGreenwalt  ASVavra  JD Chronic progressive renal disease: rate of change of serum creatinine concentration. Kidney Int. 1977;1162- 70
Link to Article
Muthusethupathi  MAKumar  SSSekaran  L Assessment of progression of chronic renal failure: using reciprocal of serum creatinine. J Assoc Physicians India. 1989;37213- 216
Kirschbaum  BB Analysis of reciprocal creatinine plots in renal failure. Am J Med Sci. 1986;291401- 404
Link to Article
Rowe  PARichardson  REBurton  PRMorgan  AGBurden  RP Analysis of reciprocal creatinine plots by two-phase linear regression. Am J Nephrol. 1989;938- 43
Link to Article
Shah  BVLevey  AS Spontaneous changes in the rate of decline in reciprocal serum creatinine. J Am Soc Nephrol. 1992;21186- 1191
Not Available, Effects of diet and antihypertensive therapy on creatinine clearance and serum creatinine concentration in the Modification of Diet in Renal Disease study. J Am Soc Nephrol. 1996;7556- 566
Buckeit  JBOlshen  RABlouch  KMyers  BD Modeling of progressive glomerular injury in humans with lupus nephritis. Am J Physiol. 1997;273 (1, pt 2) F158- F169
Schainuck  LIStriker  GECutler  REBenditt  EP Structural-functional correlations in renal disease, II: the correlations. Hum Pathol. 1970;1631- 641
Link to Article
Yarwood  SJWoodgett  JR Extracellular matrix composition determines the transcriptional response to epidermal growth factor receptor activation. Proc Natl Acad Sci U S A. 2001;984472- 4477
Link to Article
Ng  YYHuang  TPYang  WC  et al.  Tubular epithelial-myofibroblast transdifferentiation in progressive tubulointerstitial fibrosis in 5/6 nephrectomized rats. Kidney Int. 1998;54864- 876
Link to Article
Fan  JMHuang  XRNg  YY  et al.  Interleukin-1 induces epithelial-myofibroblast transdifferentiation through a transforming growth factor-β1-dependent mechanism in vitro. Am J Kidney Dis. 2001;37820- 831
Link to Article
Suzuki  TKimura  MAsano  MFujigaki  YHishida  A Role of atrophic tubules in development of interstitial fibrosis in microembolism-induced renal failure in rat. Am J Pathol. 2001;15875- 85
Link to Article
Schocklmann  HOLang  SSterzel  RB Regulation of mesangial cell proliferation. Kidney Int. 1999;561199- 1207
Link to Article
Savill  JH Regulation of glomerular cell number by apoptosis. Kidney Int. 1999;561216- 1222
Link to Article
Sugiyama  HKashihara  NMaeshima  Y  et al.  Regulation of survival and death of mesangial cells by extracellular matrix. Kidney Int. 1998;541188- 1196
Link to Article
Mooney  AJackson  KBacon  R  et al.  Type IV collagen and laminin regulate glomerular mesangial cell susceptibility to apoptosis via β(1) integrin-mediated survival signals. Am J Pathol. 1999;155599- 606
Link to Article
Sharma  KWang  RXZhang  LY  et al.  Death the Fas way: regulation and pathophysiology of CD95 and its ligand. Pharmacol Ther. 2000;88333- 347
Link to Article
Leithauser  FDhein  JMechtersheimer  G  et al.  Constitutive and induced expression of APO-1, a new member of the nerve growth factor/tumor necrosis factor receptor superfamily, in normal and neoplastic cells. Lab Invest. 1993;69415- 429
Schelling  JRCleveland  RP Involvement of Fas-dependent apoptosis in renal tubular epithelial cell deletion in chronic renal failure. Kidney Int. 1999;561313- 1316
Link to Article
Schelling  JRNkemere  NKopp  JBCleveland  RP Fas-dependent fratricidal apoptosis is a mechanism of tubular epithelial cell deletion in chronic renal failure. Lab Invest. 1998;78813- 824
Khan  SCleveland  RPKoch  CJSchelling  JR Hypoxia induces renal tubular epithelial cell apoptosis in chronic renal disease. Lab Invest. 1999;791089- 1099
Pagtalunan  MEMiller  PLJumping-Eagle  S  et al.  Podocyte loss and progressive glomerular injury in type II diabetes. J Clin Invest. 1997;99342- 348
Link to Article
Schiffer  MBitzer  MRoberts  IS  et al.  Apoptosis in podocytes induced by TGF-β and Smad7. J Clin Invest. 2001;108807- 816
Link to Article
Steffes  MWSchmidt  DMcCrery  RBasgen  JM Glomerular cell number in normal subjects and in type 1 diabetic patients. Kidney Int. 2001;592104- 2113
Link to Article
Isbel  NMHill  PAFoti  R  et al.  Tubules are the major site of M-CSF production in experimental kidney disease: correlation with local macrophage proliferation. Kidney Int. 2001;60614- 625
Link to Article
Descamps-Latscha  BHerbelin  ANguyen  AT  et al.  Balance between IL-1β , TNF-α, and their specific inhibitors in chronic renal failure and maintenance dialysis: relationships with activation markers of T cells, B cells, and monocytes. J Immunol. 1995;154882- 892
Hooke  DHGee  DCAtkins  RC Leukocyte analysis using monoclonal antibodies in human glomerulonephritis. Kidney Int. 1987;31964- 972
Link to Article
Lan  HYNikolic-Paterson  DJMu  WAtkins  RC Local macrophage proliferation in the progression of glomerular and tubulointerstitial injury in rat anti-GBM glomerulonephritis. Kidney Int. 1995;48753- 760
Link to Article
Yang  NWu  LLNikolic-Paterson  DJ  et al.  Local macrophage and myofibroblast proliferation in progressive renal injury in the rat remnant kidney. Nephrol Dial Transplant. 1998;131967- 1974
Link to Article
Nakao  ANosaka  KOhishi  N  et al.  Long-term effects of LTB4 antagonist on lipid induced renal injury. Kidney Int Suppl. 1997;63S236- S238
Kang  DJoly  AHOh  SW  et al.  Impaired angiogenesis in the remnant kidney model, I: potential role of vascular endothelial growth factor and thrombospondin-1. J Am Soc Nephrol. 2001;121434- 1437
El-Koraie  AFBaddour  NMAdam  AGEl Kashef  EHEl Nahas  AM Role of stem cell factor and mast cells in the progression of chronic glomerulonephritis. Kidney Int. 2001;60167- 172
Link to Article
Hostetter  THOlson  JLRennke  HGVenkatachalam  MABrenner  BM Hyperfiltration in remnant nephrons: a potentially adverse response to renal ablation. Am J Physiol. 1981;241F85- F93
Kato  HOsajima  AUezono  Y  et al.  Involvement of PDGF in pressure-induced mesangial cell proliferation through PKC and tyrosine kinase pathways. Am J Physiol. 1999;277F105- F112
Suda  TOsajima  ATamura  M  et al.  Pressure-induced expression of monocyte chemoattractant protein-1 through activation of MAP kinase. Kidney Int. 2001;601705- 1715
Link to Article
Mogensen  CEChristensen  CK Predicting diabetic nephropathy in insulin-dependent patients. N Engl J Med. 1984;31189- 93
Link to Article
Vora  JPDolben  JDean  JD  et al.  Renal hemodynamics in newly presenting non-insulin dependent diabetes mellitus. Kidney Int. 1992;41829- 835
Link to Article
Nelson  RGTan  MBeck  GJ  et al.  Changing glomerular filtration with progression from impaired glucose tolerance to type II diabetes mellitus. Diabetologia. 1999;4290- 93
Link to Article
Anderson  SVora  JP Current concepts of renal hemodynamics in diabetes. J Diabetes Complications. 1995;9304- 307
Link to Article
Imanishi  MYoshioka  KKonishi  Y  et al.  Glomerular hypertension as one cause of albuminuria in type II diabetic patients. Diabetologia. 1999;42999- 1005
Link to Article
Stockand  JDSansom  SC Glomerular mesangial cells: electrophysiology and regulation of contraction. Physiol Rev. 1998;78723- 744
Keller  GZimmer  GMall  GRitz  EAmann  K Nephron number in patients with primary hypertension. N Engl J Med. 2003;348101- 108
Link to Article
Najarian  JSChavers  BMMcHugh  LEMatas  AJ 20 years or more of follow-up of living kidney donors. Lancet. 1992;340807- 810
Link to Article
Kasiske  BLMa  JZLouis  TASwan  SK Long-term effects of reduced renal mass in humans. Kidney Int. 1995;48814- 819
Link to Article
Narkun-Burgess  DMNolan  CRNorman  JEPage  WFMiller  PLMeyer  TW Forty-five year follow-up after uninephrectomy. Kidney Int. 1993;431110- 1115
Link to Article
Novick  ACGephardt  GGuz  BSteinmuller  DTubbs  RR Long-term follow-up after partial removal of a solitary kidney. N Engl J Med. 1991;3251058- 1062
Link to Article
Bohle  AMuller  GAWehrmann  MMackensen-Haen  SXiao  JC Pathogenesis of chronic renal failure in the primary glomerulopathies, renal vasculopathies, and chronic interstitial nephritides. Kidney Int Suppl. 1996;54S2- S9
Bohle  Avon Gise  HMackensen-Haen  SStark-Jakob  B The obliteration of the postglomerular capillaries and its influence upon the function of both glomeruli and tubuli: functional interpretation of morphologic findings. Klin Wochenschr. 1981;591043- 1051
Link to Article
Zhu  HBunn  HF Signal transduction: how do cells sense oxygen? Science. 2001;292449- 451
Link to Article
Norman  JTClark  IMGarcia  PL Hypoxia promotes fibrogenesis in human renal fibroblasts. Kidney Int. 2000;582351- 2366
Link to Article
Ong  ACJowett  TPFirth  JDBurton  SKaret  FEFine  LG An endothelin-1 mediated autocrine growth loop involved in human renal tubular regeneration. Kidney Int. 1995;48390- 401