Nephrotic Syndrome and the Podocytopathies: Minimal Change Nephropathy, Focal Segmental Glomerulosclerosis, and Collapsing Glomerulopathy
William H. Schnaper
Jeffrey B. Kopp
The term nephrotic syndrome refers to a classic tetrad of proteinuria, hypoproteinemia, edema, and hyperlipidemia. Although a relationship among these findings was recognized as early as the 15th century, the term nephrosis first achieved widespread acceptance in the early part of the 20th century, when Volhard and Fahr employed it as one of the major divisions of bilateral kidney disease.1 Later developments, notably the advent of the percutaneous kidney biopsy, facilitated further delineation of the many forms of kidney disease that result in the nephrotic syndrome.2,3 We have divided these diseases into three general categories, as shown in Table 52.1: (1) primary nephrotic syndrome, in which the process that initiates proteinuria is not immediately apparent from histopathologic evaluation; (2) inflammatory glomerular lesions; and (3) glomerulopathy secondary to other diseases that affect the kidney. Regardless of the underlying cause, all of these diseases share a common denominator in that each involves proteinuria of sufficient severity to produce hypoproteinemia. Typically, when the serum albumin concentration falls below a critical level of approximately 2 g per deciliter, the other clinical features of the nephrotic syndrome appear.
An analysis of the diseases underlying nephrosis is complicated because studies have used varied definitions; different methods of acquiring patient populations; and groupings that reflect clinical, functional, or histologic criteria. Clearly, the relative frequency of different causes of nephrosis varies with age and has changed over time. The data in Table 52.1 were developed more than 30 years ago and indicated that approximately 80% of children with renal disease had primary nephrotic syndrome, as opposed to only 25% of adults. Chronic glomerulonephritis was responsible for about half of the cases of nephrotic syndrome in adults but only 10% to 15% of childhood cases. These glomerulonephritides may result from a systemic disease, such as systemic lupus erythematosus, or they may be idiopathic, such as in membranous nephropathy. The remaining cases of nephrotic syndrome were associated with diseases such as diabetes mellitus and amyloidosis. They accounted for up to 20% of adult cases and only a very small percentage of childhood cases. This general pattern of causes for nephrotic syndrome is still observed in most industrial countries; additional causes are more likely to be seen in developing nations.4,5,6,7,8
This chapter focuses on the group of diseases subsumed under the category of primary nephrotic syndrome. We employ this term to describe the clinical picture of nephrotic syndrome that occurs in the absence of evidence for glomerulonephritis or systemic disease that would be sufficient to account for massive proteinuria. Primary nephrotic syndrome includes patients who have been described as having minimal change nephropathy (MCN), also called lipoid nephrosis, “nil” disease, idiopathic nephrotic syndrome of childhood, minimal change nephrotic syndrome (MCNS), or steroid-sensitive or steroid-responsive nephrotic syndrome. We have chosen to use MCN as the preferred term for this entity. Nephrosis is not always present in MCN, particularly in adult patients.9,10 Although most pathologists require at least the presence of nephrotic-range proteinuria prior to treatment to make the diagnosis of MCN, occasionally, patients may lack edema, hypoalbuminemia, or hypercholesterolemia but have renal histology and ultrastructure that are otherwise typical. Furthermore, describing the lesion as a nephropathy offers a descriptor that is in parallel with the other forms of primary nephrotic syndrome, in that it is based on the defined histopathology rather than a potentially variable clinical picture. We will reserve the alternate term, MCNS, for the clinical presentation in which MCN has caused the nephrotic syndrome. Some patients with primary nephrotic syndrome have only a small amount of immunoglobulin M (IgM) deposited in the glomeruli, which usually is believed to be insignificant; some biopsy specimens reveal mild mesangial hypercellularity. These patients are considered to represent variants of MCN. A larger group of patients have significant extracellular matrix accumulation
or glomerular capillary collapse. These patients are defined as having focal segmental glomerulosclerosis (FSGS), also called focal sclerosis and hyalinosis, and collapsing glomerulopathy, respectively. A spectrum of pathologic findings may be observed. Moreover, some patients, regardless of the underlying pathology, respond to treatment with corticosteroids; others with an apparently identical histologic lesion are resistant to steroid therapy.
or glomerular capillary collapse. These patients are defined as having focal segmental glomerulosclerosis (FSGS), also called focal sclerosis and hyalinosis, and collapsing glomerulopathy, respectively. A spectrum of pathologic findings may be observed. Moreover, some patients, regardless of the underlying pathology, respond to treatment with corticosteroids; others with an apparently identical histologic lesion are resistant to steroid therapy.
TABLE 52.1 Types of Kidney Disease Causing Nephrotic Syndrome in Pediatric and Adult Patients | ||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||
An examination of the incidence of these diseases illustrates both the age dependence of diagnoses and the changing nature of the underlying lesion. In a large series of renal biopsies of 1,000 consecutive patients who presented with the nephrotic syndrome to a referral center, the relative incidence of different causes of nephrosis changed over time, with significantly more FSGS and less MCN.9 Although this study could reflect some degree of referral bias, it is likely that these numbers approximate the general distribution of histologic diagnoses in nephrotic patients. Of all children 0.5 to 19 years of age in the province of Ontario who were diagnosed as having nephrotic syndrome, the incidence of FSGS also increased over time.11 The overall proportion of patients with primary nephrotic syndrome in children was greater than 90% compared with 50% in the adult study. The increasing incidence of FSGS (which more than doubled in both children and adults) is considered further in the section on Focal Segmental Glomerulosclerosis, Collapsing Glomerulopathy, and Diffuse Mesangial Sclerosis, later in this chapter.
In addition to the clinical observation that inflammation does not underlie the proteinuria of primary nephrotic syndrome, advances in our understanding of these disorders suggest that the origin of all these diseases resides in the specialized visceral epithelial cell of the glomerular filter, the podocyte. This cell contributes to the final barrier that determines the nature of the glomerular filtrate, and podocyte lesions appear to play a critical role in progressive forms of primary nephrotic syndrome. For this reason, an increasingly accepted term for the lesions causing primary nephrotic syndrome is the podocytopathies. In Table 52.2 taxonomy of the podocytopathies is shown grouped according to histologic characteristics.12,13 An analysis of the pathology, physiology, and genetics of these diseases, and insight derived from examining acquired causes of each lesion, was used in this categorization. Patients who show few or no glomerular abnormalities by use of light microscopy include those with classical MCN and its histologic variants, and patients with Finnish-type congenital nephrotic syndrome. A second group has diffuse mesangial sclerosis and represents mostly young children with congenital forms of nephrosis, frequently resulting from a single gene mutation. The third category, FSGS, represents an entity the incidence of which is rising rapidly throughout the world. We have also chosen to define a fourth group: patients with collapsing glomerulopathy. Although this diagnosis previously has been thought to be a part of FSGS, the histologic appearance, the clinical course, and advances in our understanding of disease pathogenesis strongly suggest that it represents a distinct lesion. Further details regarding the definition of the categories in this table will be provided in this chapter.
It is important to note that the nephrotic syndrome has many physiologic consequences that are not limited to the classic tetrad of proteinuria, hypoalbuminemia, edema, and hyperlipidemia. These include abnormalities of electrolyte balance, coagulation, hormonal function, and immunity. This chapter reviews the mechanisms underlying these manifestations of nephrosis, our understanding of the pathogenesis of different forms of the podocytopathies, and the clinical features and management of patients with this entity.
PATHOPHYSIOLOGY OF THE NEPHROTIC SYNDROME
Virtually every abnormality observed in primary nephrotic syndrome can be traced directly or indirectly to the urinary loss of protein. Thus, the mechanisms responsible for this proteinuria have systemic consequences that are manifested in the clinical signs and symptoms of nephrosis.
TABLE 52.2 Classification of Primary Podocytopathies | ||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||
Mechanisms for Proteinuria
The renal factors contributing to albumin homeostasis include both glomerular filtration and tubular reabsorption. In its simplest form, the glomerulus functions as a means to promote fluid and solute flux from the blood vessel to the urinary space, from where most constituents of the filtrate are then reabsorbed. This model acquires significant complexity as solute particles approach the size limits that are characteristic of the glomerular filter. Furthermore, recent progress in understanding tubular handling of protein has demonstrated that the reabsorptive component also has a significant impact on albumin homeostasis.
Renal Handling of Macromolecules
The glomerular barrier to filtration consists of three layers: fenestrated endothelial cells, the trilaminar glomerular basement membrane (GBM), and the epithelial cell layer (Fig. 52.1). The epithelium does not constitute a continuous layer; rather, the interdigitating extensions from adjacent epithelial cells or podocytes are separated by spaces readily apparent on electron microscopy. The GBM has been considered a major barrier to filtration.14 Experimental evidence supports a hypothetical construct in which the GBM is a thixotropic gel (one containing spicules that retard the passage of macromolecules through it).15 Diffusion through this gel plays a significant role in restricting protein passage.16 Thus, the filtration of protein is restricted in the same manner that regulates protein movement during gel electrophoresis, where small molecules most easily penetrate.17 At the same time, other studies support a model in which macromolecules encounter a porous structure that limits the passage of larger molecules by steric hindrance.18 Glomerular filtration is possible because a small portion of the urinary space separates the interdigitations of the podocytes. These spaces are partly occluded by the epithelial slit diaphragm (Fig. 52.1), which has pores19,20 that likely constitute the limiting barrier structure, causing steric hindrance.21 The barrier itself is composed primarily of nephrin, a cell-cell adhesion molecule that interdigitates between adjacent cell processes22 and is supported by a slit-diaphragm complex that includes podocin, CD2-associated protein (CD2AP), FAT, Nephl, P-cadherin,
and vascular endothelial (VE)-cadherin.23 The result is a latticework of proteins with openings of approximately 4 × 14 nm.20,24 Therefore, both the GBM and the slit diaphragm contribute to steric influences on macromolecular filtration.
and vascular endothelial (VE)-cadherin.23 The result is a latticework of proteins with openings of approximately 4 × 14 nm.20,24 Therefore, both the GBM and the slit diaphragm contribute to steric influences on macromolecular filtration.
 FIGURE 52.1 The glomerular filtration barrier. The distribution of glomerular polyanion in the glomerular basement membrane and on the endothelial and epithelial cell layers is shown. LRE, lamina rara externa; LD, lamina densa; LRI, lamina rara interna; GBM, glomerular basement membrane. |
The porous component is demonstrated by permselectivity curves that plot the renal clearance of macromolecules, relative to the glomerular filtration rate (GFR), against the molecular radius, describing a sigmoid shape (Fig. 52.2) between approximately 2 and 5 nm (20 and 50 Å).25 Therefore, some restriction in the filtration of dextrans occurs with molecules of about a 2-nm radius; restriction increases with increasing molecular size and approaches 100% for molecules of a radius of 5 nm.26 In addition to size, the ability of macromolecules to cross the glomerular barrier is affected by molecular configuration, shape, deformability, and flexibility.14 Permselectivity also is modified by glomerular hemodynamic factors, although the mechanisms for these effects remain a subject of some controversy.17
Initially, it was believed that macromolecule handling could be accounted for by an isoporous model for glomerular filtration, one in which the steric hindrance of the glomerular passage of macromolecules results from the presence of uniform pores in the barrier, each with a radius of approximately 5 nm. The size of these pores may be increased in models of increased permeability of the GBM.27 However, it has become apparent that a heteroporous model may be more appropriate.28 In this model, there are two pathways: one subject to classic steric hindrance, and a “shunt” pathway unaffected by size selectivity. As demonstrated by the clearance of very large dextrans, the glomerular filtration of macromolecules through this second pathway is enhanced in most forms of nephrosis and exacerbated by colloid volume expansion,29 and is ameliorated in humans by antihypertensive therapy,30 pressor doses of angiotensin II (in contrast to the effect in rats),31 or indomethacin.32 Therefore, there appears to be a hemodynamic component to the activation of this mechanism for proteinuria. The impact of this shunt is most noticeable for large molecules (greater than 6 nm); its effect on albumin clearance remains to be determined.
Steric hindrance is not sufficient to account for all aspects of permselectivity. Although proteins are handled in a manner similar to that for inert macromolecules,14 protein clearances tend to be less than those of dextrans of comparable size.14 Part of this difference is explained by the relatively rigid structure of the proteins. However, albumin, which has an effective molecular radius of 3.6 nm, is cleared by the normal kidney considerably less than are the equivalent-sized dextran molecules. Albumin carries a negative electrostatic charge, and its clearance is only slightly less than that of similarly sized dextran molecules carrying a negative charge.33 This apparent charge selectivity has been attributed to negatively charged sialoglycoproteins in the glomerular filter,34 which are present at regularly spaced intervals in the lamina rarae of the basement membrane,35 at the endothelial fenestrae,36 and lining the epithelial podocytes.37 Collectively, these constitute the glomerular polyanions (Fig. 52.1). The presence of such negative-charge sites was proposed to be responsible for both the facilitated transport of polycations38 and the restricted transport of polyanions39 relative to that of neutral molecules of comparable size (Fig. 52.3). These effects are most apparent in the size range that is affected by some degree of steric hindrance.
Thus the determinants of glomerular permeability for a given particle are steric hindrance, glomerular hemodynamics, and electrostatic charge. The critical negative charges may not reside in the solid phase of the glomerular filter. The subpodocyte space creates a zone of delayed passage for larger macromolecules that cross the GBM.40 Within this space, streaming potential establishes a flow-dependent charge gradient that is more negative on the urinary side of the GBM. A flow-dependent electrophoretic potential is accordingly established, driving negatively charged macromolecules back toward the vascular space.41 The relative contribution of this gradient remains to be determined.
Tubular Handling of Protein
Renal protein metabolism also is affected significantly by tubular function. The glomerular filtrate normally contains a small amount of protein. A proximal tubular system has
sufficient capacity that, under physiologic conditions, little intact protein from the filtrate is present in the urine. For example, filtered albumin is subject to lysosomal degradation upon pinocytosis by the proximal tubular cell, with fragments appearing in both the plasma and the urine.42 However, studies of rat kidneys, isolated but perfused in situ with radiolabeled albumin, indicate that some albumin is reabsorbed intact.43 One mechanism of tubular protein reabsorption is demonstrated by its absence in Dent disease, a defect in chloride transport resulting from a mutation in the gene for a renal-specific, voltage-gated chloride channel, CLC-5, leading to hypercalciuric nephrolithiasis.44 Proteinuria in this disease results from disruption of both receptormediated and fluid-phase endocytosis.45 Patients with Dent disease have characteristic urinary losses of retinol-binding protein (RBP) and albumin.46 The failure of protein reabsorption in this lesion has permitted an estimate that the glomerular filtrate contains 22 to 32 mg per liter of albumin, or roughly 3 to 6 g per day in the normal human adult, virtually all of which is reabsorbed under normal conditions. This represents greater than 4% of the total plasma albumin.47
sufficient capacity that, under physiologic conditions, little intact protein from the filtrate is present in the urine. For example, filtered albumin is subject to lysosomal degradation upon pinocytosis by the proximal tubular cell, with fragments appearing in both the plasma and the urine.42 However, studies of rat kidneys, isolated but perfused in situ with radiolabeled albumin, indicate that some albumin is reabsorbed intact.43 One mechanism of tubular protein reabsorption is demonstrated by its absence in Dent disease, a defect in chloride transport resulting from a mutation in the gene for a renal-specific, voltage-gated chloride channel, CLC-5, leading to hypercalciuric nephrolithiasis.44 Proteinuria in this disease results from disruption of both receptormediated and fluid-phase endocytosis.45 Patients with Dent disease have characteristic urinary losses of retinol-binding protein (RBP) and albumin.46 The failure of protein reabsorption in this lesion has permitted an estimate that the glomerular filtrate contains 22 to 32 mg per liter of albumin, or roughly 3 to 6 g per day in the normal human adult, virtually all of which is reabsorbed under normal conditions. This represents greater than 4% of the total plasma albumin.47
 FIGURE 52.2 Permselectivity curves for patients with severe proliferative glomerulonephritis (GN) and for those with nephrotic syndrome secondary either to the minimal change nephropathy or to glomerulonephritis. Normal values are depicted by the shaded area. The arrow indicates the molecular size of albumin. The fractional clearance of larger macromolecules is increased in severe glomerulonephritis. In minimal change nephrotic syndrome (MCNS), the fractional clearance of smaller molecules is decreased. Patients with nephrotic syndrome secondary to glomerulonephritis show a hybrid curve. (Data modified from Robson AM, Cole BR. Pathologic and functional correlations in the glomerulopathies. In: Cummings NB, Michael AF, Wilson CB, eds. Immune Mechanisms in Renal Disease. New York: Plenum; 1982:109, with permission.) |
 FIGURE 52.3 Clearance of neutral dextran (D), negatively charged dextran sulfate (DS), and positively charged diethylaminoethyl (DEAF) dextran of varying molecular size in normal rats and in those made albuminuric by treatment with nephrotoxic serum (NSN). In normal animals, the clearance of negatively charged dextrans is retarded, and that of cationic dextrans is enhanced, demonstrating charge selectivity by the glomerular filter. In NSN, charge discrimination is lost. (Reprinted from Bohrer MP, Baylis C, Humes HD, et al. Permselectivity of the glomerular capillary wall: facilitated filtration of circulating polycations. J Clin Invest. 1978;61:72; by copyright permission of the American Society for Clinical Investigation.) |
This saturable mechanism for albumin reabsorption is mediated by three proteins that are associated with clathrin-coated pits in the proximal tubular cell.48 Megalin is a 600-kDa, transmembrane protein and a member of the low-density lipoprotein-receptor family. It colocalizes in cultured opossum kidney (OK) cells with exogenous albumin and with cubilin, a 460-kDa protein that does not have a transmembrane domain. A third protein, amnionless, forms a complex with megalin and cubilin; the three proteins collaborate to reabsorb albumin.49 Ligands for cubilin in the glomerular filtrate include not only albumin but also immunoglobulin light chain and apolipoprotein (apo)A-I. Megalin binds to the vitamin-binding proteins, RBP and vitamin D-binding protein, hormones, enzymes and β2– and α1– microglobulin, as well as albumin.50,51 As will be discussed in the section to follow on Consequences of Proteinuria, the loss of many of these proteins has clinical significance in nephrotic syndrome. Another albumin “rescue” pathway that appears to facilitate reabsorption of intact albumin has been attributed to the FcRn immunoglobulin receptor.52
Altered Permselectivity in Nephrosis
Permselectivity patterns obtained in patients with MCNS53,54 (Fig. 52.2) or animal models of selective albuminuria39 (Fig. 52.3) show a relative decrease in macromolecular clearance even in the presence of marked proteinuria. In contrast, patients with glomerulonephritis show increased macromolecular clearances (Fig. 52.2), presumably due to structural damage to the GBM, which may be visible in renal biopsy material from patients in these disease states. This concept is supported by work in animals.55 Thus the mechanisms for proteinuria in MCN and FSGS appear to be distinct from those in glomerulonephritis. In the former, proteinuria is relatively selective for albumin and occurs even though clearance of macromolecules comparable in size to albumin is decreased. In the latter, permselectivity of macromolecules that are 2.5 nm (25 Å) or larger is increased, resulting in poorly selective proteinuria. Patients with glomerulonephritis who have proteinuria that is sufficiently severe to cause the nephrotic syndrome may show a pattern of permselectivity (Fig. 52.2) that is a hybrid between those found in MCNS and those found in uncomplicated glomerulonephritis.53 In these patients, as in MCNS, clearance of smaller molecules is relatively decreased. However, in contrast to the situation in MCNS, the relative clearance of larger molecules is increased. Similar hybrid curves have been described in diabetic glomerulosclerosis.56
Therefore, nephrotic proteinuria does not result from a simple defect in glomerular filter steric hindrance. Several theories have been advanced to account for albumin loss. A prominent one is a decrease in glomerular electrostatic charge selectivity. Renal biopsy material from patients with nephrotic syndrome shows decreased staining for glomerular polyanion.57,58,59,60 Indeed, studies in MCNS patients suggested that albuminuria results from a reduction of fixed negative charge by approximately 50%.61 Rats with nephrotic syndrome induced by puromycin amino nucleoside (PAN), which causes predominant albuminuria, show decreased staining by cationic dyes62 and decreased sialic acid content.63 Animals with PAN-induced nephrotic syndrome64 as well as those with acute heterologous nephrotoxic serum nephritis33 show increased clearance of negatively charged dextrans, with permselectivity curves approximating those of neutral dextrans. Further, the intravenous infusion of various polycations into animals results in a loss of staining for glomerular polyanion, increased porosity of the glomerular filter, and heavy proteinuria.65,66,67 Unilateral renal artery infusion of the polycation protamine sulfate causes ipsilateral albuminuria and depletion of glomerular polyanion.67 Finally, studies in patients suggest that the neutralization of vascular anionic charges may be systemic in nature68,69 rather than confined to the kidney. This could result from effects of a protease present in the circulation such as hemopexin.70,71 However, sieving curves generated in rats by glomerular localization of neutral or negatively charged polysaccharides were unable to demonstrate charge selectivity of the glomerular filter,72 suggesting that technical factors or differences in the experimental approach could significantly affect the validity of experiments demonstrating charge selectivity. In these studies, bovine serum albumin (BSA) “uptake” was extremely high relative to other markers. A recent modification of the charge-selective model was proposed by Hausmann and colleagues.41 In this flow-dependent model involving an electrochemical gradient across the filtration barrier, podocyte effacement disturbs flow, decreasing the negative charge in the urinary space. Consistent with this model, in PAN nephrosis increased albumin permeability occurs only in areas of podocyte dysfunction rather than diffusely, suggesting that the podocyte is more critical than the slit diaphragm in the pathogenesis of albuminuria.73
Another study found a role for negative charge in modulating renal protein handling in rats infused with neutral or anionic horseradish peroxidase. These results suggested that proteins may be more affected than polysaccharides by charge, but in this model charge selectivity was lost after inhibiting tubular protein uptake with lysine or ammonium chloride, suggesting the conclusion that charge selectivity does not reside in the glomerulus.74 In mice, the plasma elimination rate of albumin (effective molecular radius of 36 Å) was comparable to that of much larger Ficoll molecules (≥ 65 Å). When the animals were treated with PAN, albumin clearance increased through an unknown renal mechanism.75 Although charge selectivity could explain these findings, any mechanism that is involved could have affected either the glomerulus or the tubule. A study in analbuminemic rats treated with PAN showed no effect on renal size selectivity with treatment. There also was no change in the characteristics of urinary protein excretion.76 Previously, it has been assumed that a greatly increased delivery of albumin to the proximal tubule saturated transport mechanisms or that the other proteins lost in the urine were bound to albumin, in either case causing loss of the nonalbumin proteins. However, the pattern of proteinuria observed in this study was similar to the findings in the CLC-5-null mouse, where low-molecular-weight proteinuria occurs, and suggests that at least some protein losses in nephrosis result from specific tubular mechanisms. These data support a significant role for the derangement of tubular protein handling in nephrotic syndrome.
Several investigators have suggested that impairment of the rescue pathway and other tubular, rather than glomerular, mechanisms are a significant cause of nephrotic proteinuria.17,77 However, several lines of evidence suggest that this is not the case. Patients with Dent disease, and the CLC-5-null mouse, have “nephrotic-range” proteinuria but do not have nephrosis. The salvage of some of the 4% of the plasma albumin that is filtered per day47 is likely to contribute positively to homeostasis, but a significant portion of albumin rescued by the tubule is degraded.42 Assuming that half of the albumin is reclaimed intact, it is unlikely that losing 2% of plasma albumin per day will have a major effect on plasma albumin concentration. Finally, Deen and Lazzara78 modeled the sieving coefficient for albumin. They performed a mass-transfer analysis to determine whether the sieving coefficient could be similar to, rather than greatly less than, that for neutral Ficoll of the same size. The higher value, which would have been required in the models supported by adherents of a causal role for tubular proteinuria, would generate tubular albumin concentrations located 1 mm distal to the glomerulus that are 20-fold higher than has been measured by rat micropuncture studies. The authors concluded that the glomerulus was the primary restricting site for albuminuria.78 Although the possibility of charge selectivity was considered, it could not be tested by this analysis.
The alteration of podocyte architecture may account for the generally decreased fractional clearance of smaller macromolecules in nephrotic syndrome cited previously. It has been suggested that simplification of the foot process makes the glomerular pore less complex, thereby allowing for increased clearance of some long, narrow, rigid molecules. However, most plasma proteins are prolate ellipsoids (stubby cigar shaped) and show decreased clearance.79 The effective pore radius was reported to be decreased in both MCNS and FSGS.80 In this study, the ratio of total pore area to pore length was reduced by more than 50%. Further support for decreased pore area is found in studies indicating decreased filtration slit frequency (likely secondary to foot-process fusion)81 or decreased pore number.82 These findings would account for decreased macromolecular clearance but not enhanced albumin clearance.
An alternative to charge neutralization as an explanation of proteinuria is suggested by the data indicating that permselectivity patterns show enhanced clearance of larger neutral dextrans in FSGS.82 Studies by Yoshioka and colleagues83 suggest that there is enhanced clearance of albumin by less affected glomeruli, implicating hemodynamic factors related to hyperfiltration in remnant nephrons.84 This could be accounted for by the heteroporous model in which a different class of pores greater than 60 Å in radius is increasingly used. This shunt pathway is active in angiotensin II-stimulated proteinuria85 and Heymann nephritis.86 Further, the hemodynamic implications of proposing a role for the shunt pathway are supported by the salient effect of angiotensin-converting enzyme (ACE) inhibition on glomerular size selectivity in disease.87 However, because patients with MCN do not necessarily have increased use of the shunt pathway,82 it is not clear that shunting represents a mechanism of proteinuria common to all causes of the nephrotic syndrome or is the major cause of nephrotic albuminuria.
A third hypothesis regarding the stimulus for proteinuria could account for changes in both charge and steric hindrance in the glomerular filtration barrier. Small anions such as Cl– are freely filtered by the glomerulus, whereas negatively charged molecules such as albumin that are large enough to interact with the filtration barrier (but pass through) are affected by electrostatic hindrance. Modest increases in slit diaphragm pore size, perhaps mediated by alterations in cytoskeletal function,88 may be sufficient to both decrease the steric hindrance of and reduce electrostatic interference with albumin transit, even before considering the amplifying effects of barrier charge neutralization or increased shunt pathway use.
Molecular Mechanisms of Podocyte Effacement
Given the importance of podocyte architecture in all of these models, the regulation of podocyte shape is an essential determinant of proteinuria. In podocyte effacement, the normally cortical distribution of the actin cytoskeleton, where structures extend from the periphery into the individual foot processes, is disrupted. Mundel and colleagues89 have described a model system wherein the molecule B7-1 (also known as CD80), previously associated with lymphocyte
costimulatory signals, is expressed in injured podocytes and binds to cytoskeletal structures, causing foot-process effacement.89,90 It is expressed in the glomeruli of patients with MCN in relapse, but not in remission, and only marginally in patients with FSGS,91 suggesting one mechanism by which podocyte architecture could be disrupted. In several mouse models of proteinuria, circulating levels of soluble urokinase receptor (suPAR) interfere with podocyte adhesive interactions through the adhesion molecules β3-integrin,92 presumably disrupting podocyte architecture through this mechanism as well.
costimulatory signals, is expressed in injured podocytes and binds to cytoskeletal structures, causing foot-process effacement.89,90 It is expressed in the glomeruli of patients with MCN in relapse, but not in remission, and only marginally in patients with FSGS,91 suggesting one mechanism by which podocyte architecture could be disrupted. In several mouse models of proteinuria, circulating levels of soluble urokinase receptor (suPAR) interfere with podocyte adhesive interactions through the adhesion molecules β3-integrin,92 presumably disrupting podocyte architecture through this mechanism as well.
Consequences of Nephrotic Proteinuria
It is generally accepted that the central feature of the nephrotic syndrome, irrespective of its underlying renal cause, is hypoalbuminemia resulting from urinary loss. There is increased fractional catabolism of albumin in nephrotic syndrome,93 mostly within the renal tubule after the increased filtration of plasma proteins.94 Rates of hepatic synthesis of albumin are increased,95 but this increase is inadequate to compensate for urinary losses.96 Although gastrointestinal losses are possible through the transudation of albumin across the bowel wall in nephrosis, these are not likely to contribute significantly to decreased plasma albumin concentrations in the absence of significant bowel pathology.
Nonetheless, urinary losses cannot be considered an isolated phenomenon. It is apparent that a special relationship exists among protein synthetic capability, urinary loss of protein, and plasma protein concentrations. For example, patients undergoing chronic peritoneal dialysis lose “nephrotic range” amounts of protein, yet they usually have close to normal serum albumin concentrations.97 In nephrosis, the rate of hepatic albumin synthesis is related to dietary protein intake. However, increasing protein intake leads to glomerular hyperfiltration98,99 and enhanced loss of protein in the urine, resulting in lower serum albumin concentrations in patients on high protein diets.100 The increase in dietary intake appears to stimulate selective hepatic expression of messenger ribonucleic acid (mRNA) for albumin, indicating that the stimulus is specific for albumin production and is not generalized to other proteins as well.101 The dietary stimulus can be dissociated from potential effects of alterations in plasma oncotic pressure.102 Although specific plasma amino acid content is unchanged, nitrogen balance is rendered more positive by ACE inhibition,103 which decreases hyperfiltration and thus the amount of protein lost in the urine. Indeed, enalapril decreases UAlbumin V (absolute albumin excretion) and fractional catabolism of albumin in normal or nephrotic rats on high protein diets.104,105
 FIGURE 52.4 The forces that govern the movement of fluid across the peripheral capillary wall in healthy persons and in patients with primary nephrotic syndrome. The shaded area represents the lumina of the capillaries. The size and direction of the arrows are in proportion to the magnitude and direction of the force described by that arrow. In minimal change nephrotic syndrome, hypoalbuminemia causes a marked reduction in oncotic pressure. This increases the driving force for fluid out of the arteriolar end of the capillary and decreases the forces available for return of fluid at the venous end. The result is the development of increased amounts of fluid in the interstitial space and the beginning of edema formation. See text for more details. (From Robson AM. Edema and edema forming states. In: Klahr S, ed. The Kidney and Body Fluids in Health and Disease. New York: Plenum; 1984:119, with permission.) |
Edema Formation
One of the major consequences of hypoalbuminemia is edema formation. The major forces that maintain vascular volume are believed to be those described by Starling,106 namely, the algebraic sum of hydrostatic and oncotic pressures acting at the level of the peripheral capillary beds (Fig. 52.4). Hydrostatic pressure is the dominant force at the arteriolar end of the capillary, where it is generated by
arterial blood pressure. Pressure is lower in the capillaries (40 to 45 mm Hg) than in the arterial system, but it is markedly higher than tissue pressure, which ranges from 2 to 5 mm Hg. Hydrostatic pressure is opposed by plasma oncotic pressure (the osmotic pressure generated by colloidal solute), which is 25 to 30 mm Hg in healthy individuals. The resulting net force (10 to 15 mm Hg) drives an ultrafiltrate of blood from the capillaries into the interstitial fluid space. By the venous end of the capillary, hydrostatic pressure has been further dissipated (Fig. 52.4) and is exceeded by oncotic pressure, so that there is a net force for the return of fluid into the capillaries. In health, the loss of fluid at the arteriolar end of the capillaries slightly exceeds the amount resorbed at the venous end. The difference is returned to the circulation through the lymphatic system.107
arterial blood pressure. Pressure is lower in the capillaries (40 to 45 mm Hg) than in the arterial system, but it is markedly higher than tissue pressure, which ranges from 2 to 5 mm Hg. Hydrostatic pressure is opposed by plasma oncotic pressure (the osmotic pressure generated by colloidal solute), which is 25 to 30 mm Hg in healthy individuals. The resulting net force (10 to 15 mm Hg) drives an ultrafiltrate of blood from the capillaries into the interstitial fluid space. By the venous end of the capillary, hydrostatic pressure has been further dissipated (Fig. 52.4) and is exceeded by oncotic pressure, so that there is a net force for the return of fluid into the capillaries. In health, the loss of fluid at the arteriolar end of the capillaries slightly exceeds the amount resorbed at the venous end. The difference is returned to the circulation through the lymphatic system.107
Albumin, because of its abundance and its relatively small molecular size, is the plasma protein primarily responsible for the generation of oncotic pressure.108 A decrease in plasma albumin concentration thus results in a decrease in oncotic pressure, so that the net driving force for loss of fluid at the arteriolar end of the capillary bed is increased and that for return of fluid at the venous end is reduced. Consequently, fluid accumulates in the interstitial space, initiating edema formation. This accumulation occurs first where tissue pressure is lowest, for example, in the eyelids or in the scrotum; it also appears in the most dependent parts of the body because venous hydrostatic pressure is highest at these sites and is transmitted to the venous end of the capillaries.
In this traditional model of nephrotic edema formation, often referred to as “underfilling,”109 the translocation of fluid from the vascular to the interstitial fluid space as edema forms should decrease blood volume. The physiologic responses precipitated by such a reduction would then be important factors in producing the massive amounts of edema often seen in nephrotic syndrome. These changes include the release of antidiuretic hormone (ADH), the release of renin with increased production of angiotensin II, and decreases in renal blood flow and GFR.110,111,112 All these changes favor renal retention and positive balances of both sodium and water unless intakes are decreased. Indeed, patients may exhibit increased thirst, which is probably stimulated both by angiotensin II113 and by the decrease in blood volume monitored through baroreceptors and volume receptors. Retained sodium and water do not remain in the vascular space. Because of the hypoalbuminemia, they add to the edema.
In practice, the pathophysiology of edema formation in nephrotic syndrome is more complex than this traditional concept. Animal studies have documented that hypoproteinemia alone does not result in edema.114 Humans with congenital analbuminemia do not develop nephrosislike edema and have a normal plasma volume even in the virtual absence of serum albumin.115 Furthermore, if the traditional theory is correct, patients in relapse of nephrotic syndrome should have decreased blood volumes and values should return to normal during remission from the disease. Although reduced values for blood volume have been reported,116 normal or even increased levels have been documented too.117,118 A survey of the literature119 found that only 38% of patients with nephrosis had measurements indicating blood volumes reduced by 10% or more from normal, 48% had normal values, and 14% had increased values. In addition, patients with carefully documented MCNS studied during relapse and again during remission did not show a consistent increase in blood volumes with remission; indeed, in most, the values did not change.119,120 Therefore, in contrast to the “underfilling” model, others have proposed an “overflow” hypothesis, in which the vascular tree is filled to excess, with increased hydrostatic pressure leading to fluid extravasation. Remarkably, nail bed micropuncture measurements in nephrotic patients did not support the notion that capillary overfilling occurs, but capillary leak appeared more important than underfilling in differentiating nephrotic from normal subjects.121
There are several possible explanations for these conflicting models, each of which appears valid in some cases. One is that the reported patients had varying underlying causes of their nephrotic syndrome, in some cases involving significantly decreased renal function. Nephrotic qsyndrome secondary to glomerulonephritis usually is associated with a normal or expanded blood volume.122 A second potential confounding factor is that some patients were receiving treatment when studied. In addition to specific treatments, albumin infusion might enhance volume, and diuretic therapy may reduce both blood and interstitial fluid volume in nephrotic subjects.123 A third issue is that measurements of blood volume are difficult to interpret because of methodologic problems. Labeled red cells may not circulate ideally in volume-depleted states, so that peripheral hematocrit may not reflect total body hematocrit; labeled albumin may have an increased volume of distribution in nephrotic syndrome, especially if vascular integrity to albumin is decreased.124 Thus, both methods could be subject to errors.122 Indeed, if the suggestion that nephrotic syndrome involves a generalized decrease of negative-charge sites69 is correct, loss of such charge sites in capillary beds could cause increased losses of albumin into edema fluid.125 This transfer not only would alter the apparent volume of distribution for albumin, but also might increase net extravascular oncotic pressure at the level of the capillaries. Support for this hypothesis is found in the observation that large changes in extracellular fluid volume cause little change in plasma volume in nephrotic patients.126
An attractive explanation for variations in reported blood volume is that the patients were studied in different phases of their disease process.127 Blood volume could be reduced during the pathogenesis of the nephrotic state, particularly in MCNS, but return to normal as anasarca develops. The decrease in plasma volume after experimental depletion of serum proteins can be prevented by massive expansion of the extracellular fluid with saline solution.128 Nephrotic subjects progress through a sodium-retaining phase but
eventually enter into a new steady state in which they no longer accumulate edema and once again demonstrate the ability to excrete a sodium load.111 With this new steady state, sodium and water retention may be so marked and edema accumulation may be so massive that tissue hydrostatic pressure is increased and blood volume is returned to normal. This may explain reports in which nephrotic subjects could be separated into those with high and those with low urine sodium concentrations.120 The high-volume state, whether from massive fluid intake or decreased renal function, represents the overflow pathogenesis of nephrotic edema. It is likely that both underfilling and overflow occur, perhaps at different times in the same patient.
eventually enter into a new steady state in which they no longer accumulate edema and once again demonstrate the ability to excrete a sodium load.111 With this new steady state, sodium and water retention may be so marked and edema accumulation may be so massive that tissue hydrostatic pressure is increased and blood volume is returned to normal. This may explain reports in which nephrotic subjects could be separated into those with high and those with low urine sodium concentrations.120 The high-volume state, whether from massive fluid intake or decreased renal function, represents the overflow pathogenesis of nephrotic edema. It is likely that both underfilling and overflow occur, perhaps at different times in the same patient.
Hormonal Mechanisms
Regardless of whether underfilling or overflow is paramount, a third model suggests that hormonal mechanisms are of primary importance. Although we have emphasized a primary role for hypoalbuminemia in oliguria and sodium retention, patients with MCNS often undergo a marked, remission-induced diuresis beginning as soon as urinary albumin concentrations start to decrease and before the normalization of serum albumin. Initial studies of the pathophysiology of nephrosis suggested that fluid redistribution results in aldosterone-mediated sodium retention designed to replenish vascular volume129,130 Accordingly, aldosterone activity was thought to be more important in the genesis of fluid retention than either serum albumin or colloid osmotic pressure.120 Consistent with this notion, patients with nephrotic syndrome show an increase in distal renal tubular sodium reabsorption.111,131 Increased tubular sensitivity to aldosterone may further enhance edema formation.132
Renin-Angiotensin-Aldosterone System
Inconsistencies in reported plasma renin activity (PRA) results could be due to clinical factors similar to those that confound the interpretation of blood volume measurement. These include different stages of both disease process and sodium balance, as well as variations in therapeutic regimen. For example, immunofluorescence staining of reninproducing cells in renal biopsy material from nephrotic patients revealed increased numbers of these cells in hypoalbuminemic states. However, the increase correlated with a number of variables, most notably the presence of vascular disease.133 In an attempt to standardize some of these variables, renin-sodium profiles were performed on patients with nephrotic syndrome. Two groups of patients were identified. In keeping with traditional concepts, the classic form was typically seen in patients with MCNS, in whom high levels of PRA and aldosterone activity were associated with vasoconstriction and hypoalbuminemia; values were further stimulated rather than suppressed by salt loading and decreased spontaneously before the occurrence of steroidinduced diuresis. In the hypervolemic, overfilling, form, seen typically with chronic glomerulonephritis and renal insufficiency, low renin activity was associated with sodium retention and increased normally with sodium depletion.122 Other studies correlated PRA with plasma volume, serum albumin concentration,134 or the state of sodium balance.119 Natriuresis in MCNS was associated with an increase in PRA and presumably a decrease in plasma volume,119 whereas that induced by water immersion, presumably mediated by an increase in blood volume, was associated with a measured decrease in PRA.135 Therefore, PRA appears to correlate better with plasma volume than with the rate of urinary sodium excretion.
Difficulties in confirming a definitive role for the renin-angiotensin system in the genesis of nephrotic edema are similar to those in explaining edema formation in cirrhosis.136 A multiplicity of interacting factors may be responsible in both of these disease states. Therefore, plasma renin levels could be controlled tightly by a variety of feedback mechanisms so that subtle changes, too small to be detected by current laboratory methods, are all that occur to maintain the altered homeostasis.
It was proposed that aldosterone could mediate sodium retention through its actions on the epithelial sodium channel, ENaC. In support of this hypothesis, ENaC shows increased apical targeting in nephrotic rats.137 However, although adrenalectomy prevents this apical targeting, the adrenalectomized rats continue to show significant sodium retention and edema.138 While these results do not rule out an alternative mechanism of aldosterone action, they implicate nonaldosterone mechanisms in nephrotic sodium retention.
Other Hormonal Regulators of Fluid and Electrolyte Balance
Other factors affecting volume status may include abnormal vascular tone,139 altered levels of catecholamines,140 and a variety of hormonal mechanisms.
Antidiuretic hormone secretion. ‘Nephrotic patients with MCN may show decreased solute-free water excretion, although the capacity to generate solute-free water remains intact.112 Increased ADH secretion may reflect a physiologic response to decreased intravascular volume.116 In contrast, maximal urine osmolarity may be decreased in experimental rat nephrosis due to decreased renal tubular expression of aquaporin.141,142
Prostaglandin metabolism. Elevated levels of prostaglandin E2 (PGE2) were found in the serum of patients with nephrotic syndrome, the majority of whom had MCN.143 The highest values were observed when the patients had clinically apparent edema. Urinary PGE2 levels were increased in patients with idiopathic nephrotic syndrome who had a low urine sodium concentration as well as elevated plasma renin-aldosterone activity.144 The observation that the administration of indomethacin to nephrotic patients results in
an increase in body weight and a decrease in GFR suggests that prostaglandins may play a role in either the maintenance of GFR or amelioration of edema in nephrotic syndrome. Indomethacin also decreased proteinuria and PRA.144 Response to indomethacin is dependent on concurrent sodium intake. When the agent was given to nephrotic patients on sodium-restricted diets, it resulted in a decrease in GFR. A similar drug regimen for patients with more liberal sodium intake did not affect renal hemodynamics.145
an increase in body weight and a decrease in GFR suggests that prostaglandins may play a role in either the maintenance of GFR or amelioration of edema in nephrotic syndrome. Indomethacin also decreased proteinuria and PRA.144 Response to indomethacin is dependent on concurrent sodium intake. When the agent was given to nephrotic patients on sodium-restricted diets, it resulted in a decrease in GFR. A similar drug regimen for patients with more liberal sodium intake did not affect renal hemodynamics.145
Atrial natriuretic peptide. Because atrial natriuretic peptide (ANP) causes renal vasodilation, an increase in GFR, and increased sodium excretion,146 it has been suggested that abnormal metabolism of this hormone could mediate sodium retention in nephrosis. The acute increase of plasma volume following albumin infusion in nephrotic children is accompanied by a fivefold increase in ANP levels.147 However, this may simply reflect a change from low plasma volume status before the infusion is begun in patients who likely have MCNS. Plasma concentrations of ANP were determined to be low in nephrotic patients compared to patients who had acute glomerulonephritis, and ANP levels correlated well with the degree of edema in nephritis but not in nephrosis.148 Therefore, regulation of ANP appeared to be appropriate for presumed volume status. In rats with Adriamycin-induced nephrotic syndrome, changes in GFR after the infusion of ANP were similar to those in control animals, indicating that nephrosis does not alter glomerular filtration by changing ANP sensitivity.149 In a similar model, no change was detected in ANP receptor density in nephrotic kidneys.150 Nephrotic patients respond physiologically to ANP infusion,151 although the mechanism by which this occurs may be different from that in normal subjects.152 It has been proposed that ANP mediates the diuretic response to head-out immersion in nephrosis,153 but the effect of ANP infusion, unlike that of immersion, is blocked by enalapril.154
Physical and Anatomic Factors Affecting Glomerular Filtration Rate
Taken together, these findings suggest that, although the secretion of ANP may in part mediate diuresis, physical factors are of greatest importance in the fluid retention of nephrosis, with abnormalities of ANP representing appropriate responses for the patient’s physiology.155 These physical factors may include a significant intrarenal component. Children with MCNS have decreases in both GFR and filtration fraction.110,156,157 Decreased GFR could be due to a decrease in the ultrafiltration coefficient (Kf), causing a reduction in single-nephron GFR,155 and has been suggested to result from effacement of the glomerular epithelial cell foot processes.158 Alternatively, the decreased GFR could be a consequence of raised intratubular hydrostatic pressure in the proximal tubule secondary to the presence of filtered albumin, an increase in resistance to tubular flow,159 or decreased proximal reabsorption of tubular fluid as a result of a reduction in peritubular capillary oncotic pressure.155 There also may be a local role for the renin-angiotensin system, as saralasin infusion in experimental unilateral PAN-induced nephrotic syndrome resulted in an increase in single-nephron GFR in the experimental, but not in the control, kidney.155 In another animal model of nephrotic syndrome, that of nephrotoxic serum nephritis, the Kf was reduced, but compensatory mechanisms maintained renal blood flow and whole-kidney and single-nephron GFR. These responses appeared to be intrarenal in origin and caused an increase in glomerular capillary pressure.160
Another factor affecting GFR is plasma albumin concentration. Hypoalbuminemia has been postulated to decrease glomerular plasma flow, thereby decreasing GFR. However, lower albumin also decreases plasma oncotic pressure, which should increase GFR. Löwenborg and Berg161 report that GFR and filtration fraction vary directly with serum albumin but vary inversely with mean arterial blood pressure in children with MCNS. This finding supports a role for altered Kf in relapse, which is consistent with foot-process effacement. Kf is determined by the total filtration slit length, as shown by mathematical modeling of experimental data.81 Some studies suggest that there is a weak correlation between the amount of proteinuria and the extent of foot-process effacement, but these studies included patients in relapse and in remission with MCN162 and with multiple nephrotic diseases.163 A study of 23 MCN patients in relapse showed no correlation between proteinuria and foot-process effacement (r = 0.25, P = .25).164 The authors make the important point that assessment of the quantitative relationship between podocyte foot-process effacement and proteinuria should exclude patients in remission, as these patients have normal podocyte morphology (and therefore do not address the hypothesis that the degree of effacement correlates with proteinuria). Foot-process effacement reverses when patients undergo spontaneous or glucocorticoid-induced remission.
Other Physiologic Changes in Fluid and Electrolyte Metabolism
A curious phenomenon in primary nephrotic syndrome, perhaps related to decreased filtration fraction, is the occurrence of reversible or permanent renal failure unexplained by the underlying disease process. This has been reported in association with both MCN165,166,167 and FSGS.139 In some patients, renal failure was associated with the use of nonsteroidal anti-inflammatory drug (NSAID) therapy.168,169 These episodes occur in the absence of renal vein thrombosis (vide infra) or other systemic symptoms. Because fractional excretion of sodium is low in these patients,170 it is likely that the marked decrease in GFR occurs for hemodynamic reasons171 rather than because of acute tubular necrosis or vasomotor nephropathy. In a study of 15 patients with MCNS and renal failure, GFR measured by inulin clearance was decreased out of proportion to clearance of para-aminohippurate (PAH), with filtration fraction reduced to between 3% and 9%.172
Improvement of renal function occurred in association with diuretic therapy either with or without albumin infusion. In patients who improved with pharmacologic diuresis, the serum creatinine level again rose on return to an edematous state. The authors postulate that glomerular hemodynamics were altered by the presence of intrarenal edema, which occurred concomitantly with peripheral edema.
Improvement of renal function occurred in association with diuretic therapy either with or without albumin infusion. In patients who improved with pharmacologic diuresis, the serum creatinine level again rose on return to an edematous state. The authors postulate that glomerular hemodynamics were altered by the presence of intrarenal edema, which occurred concomitantly with peripheral edema.
Other circulatory abnormalities have been observed in patients with nephrotic syndrome. The occurrence of hypovolemic shock and hypotension has been related to a variety of medical procedures.173 However, hypotension may occur spontaneously. These episodes usually are seen in patients during relapse who have an intercurrent illness causing fluid loss, such as emesis or diarrhea. The patients usually show marked responsiveness to small amounts of intravenous saline that are insufficient to replenish all fluid losses, suggesting a failure in maintenance of vascular tone. Recovery usually occurs if this complication is identified early and treated promptly. Sequelae may include acute tubular necrosis, renal vein thrombosis (RVT), or death.
Hyperlipidemia
Lipemic serum has long been recognized as a cardinal feature of the nephrotic syndrome.174 Abnormalities in postprandial lipid metabolism were described more than 40 years ago.175 Biochemical evaluation has shown that all lipid components of the plasma are increased, with cholesterol increasing more rapidly than phospholipid. Thus, as acute severity of the disease worsens, as measured by proteinuria, increased lipid levels176 and the ratio of cholesterol to phospholipid increases.177 Triglycerides are relatively normal at the initiation of relapse but increase as the disease continues178; lactescence occurs when the plasma triglyceride content exceeds 400 mg per deciliter. Hyperlipidemia may persist well into remission,179 suggesting a residual effect of nephrosis on lipoprotein transport.180
 FIGURE 52.5 Normal pathways of lipid metabolism. apo–CII, apolipoprotein C-II; Apo-E R, chylomicron remnant (apo E) receptor; B-100 R, apolipoprotein B-100 (LDL) receptor; HDL, high-density lipoprotein; IDL, intermediate-density lipoprotein; HMG CoA reductase, 3-hydroxy-3-methylglutaryl coenzyme A reductase; LCAT, lecithin-cholesterol acyltransferase; LDL, low-density lipoprotein; LPL, lipoprotein lipase; VLDL, very-low-density lipoprotein. Figure composed with the assistance of Nader Rifai. |
Depending on the classification employed, the most common patterns of hyperlipoproteinemia seen in nephrosis are types II and IV181 or types IIa, IIb, and V.182 Low-density lipoproteins (LDLs) and very low-density lipoproteins (VLDLs) show the greatest increase in concentration. Values for high-density lipoprotein (HDL) cholesterol have been reported to be elevated,183,184 normal,185,186 or decreased.178,187,188 This variation may relate to the age of the patients studied, the underlying cause of the nephrotic syndrome, the patient treatment, and whether renal insufficiency is present. Studies of lipoprotein cholesterol have produced conflicting results. The ratio of cholesterol to phospholipids or to triglycerides in various lipoproteins is altered, indicating abnormalities in quality as well as quantity of lipoproteins.
Several events may contribute to these abnormalities. Lipid metabolism is normally accomplished through a series of complex steps (Fig. 52.5). Through the action of 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA)
reductase, mevalonate is produced from acetate in the liver. This in turn is used to make cholesterol, which is incorporated into lipoproteins. The greater the triglyceride content of the lipoprotein, the less dense it is. Dietary fat absorbed from the intestine is formed into chylomicrons by being surrounded with a coat of apolipoprotein (apo) that is critical for transport of the hydrophobic lipid. The triglyceride content of the chylomicron is reduced in the periphery (mainly by the action of lipoprotein lipase [LPL]), and the resulting particle containing apo B-48 and apo E binds to the hepatocyte via a chylomicron remnant receptor. VLDL is synthesized in the liver and metabolized in the periphery through the action of LPL to intermediate-density lipoprotein (IDL), and then to LDL. LDL is bound to apo B-100, which is then taken up by the hepatocyte LDL receptor.189,190 This brings additional cholesterol back to the liver, suppressing HMG CoA reductase activity and decreasing new cholesterol synthesis. The liver also produces HDL, which participates as a transport protein in the catabolism of lower density moieties, being regenerated by lecithin-cholesterol acyltransferase (LCAT). HDL also carries apo C-II, which activates LPL. Abnormalities at any step of metabolism from lipid uptake to the enterohepatic secretion of bile could result in the hyperlipidemia of nephrosis. Likely contributing factors include increased hepatic synthesis of lipoprotein, abnormal transport of lipid through the metabolic pathway, and abnormal catabolism secondary to decreased enzyme activity.
reductase, mevalonate is produced from acetate in the liver. This in turn is used to make cholesterol, which is incorporated into lipoproteins. The greater the triglyceride content of the lipoprotein, the less dense it is. Dietary fat absorbed from the intestine is formed into chylomicrons by being surrounded with a coat of apolipoprotein (apo) that is critical for transport of the hydrophobic lipid. The triglyceride content of the chylomicron is reduced in the periphery (mainly by the action of lipoprotein lipase [LPL]), and the resulting particle containing apo B-48 and apo E binds to the hepatocyte via a chylomicron remnant receptor. VLDL is synthesized in the liver and metabolized in the periphery through the action of LPL to intermediate-density lipoprotein (IDL), and then to LDL. LDL is bound to apo B-100, which is then taken up by the hepatocyte LDL receptor.189,190 This brings additional cholesterol back to the liver, suppressing HMG CoA reductase activity and decreasing new cholesterol synthesis. The liver also produces HDL, which participates as a transport protein in the catabolism of lower density moieties, being regenerated by lecithin-cholesterol acyltransferase (LCAT). HDL also carries apo C-II, which activates LPL. Abnormalities at any step of metabolism from lipid uptake to the enterohepatic secretion of bile could result in the hyperlipidemia of nephrosis. Likely contributing factors include increased hepatic synthesis of lipoprotein, abnormal transport of lipid through the metabolic pathway, and abnormal catabolism secondary to decreased enzyme activity.
Lipoprotein Synthesis
It is clear that the hepatic synthesis of lipoproteins is increased in nephrotic patients.177,191,192,193,194 The signal for this event appears to be related to hypoalbuminemia because the daily infusion of albumin into nephrotic patients, sufficient to raise serum levels, also decreases serum lipid, triglyceride, and cholesterol levels.195 Increasing the plasma oncotic pressure in nephrotic patients or animals, by infusion of dextrans, decreases hepatic lipoprotein synthesis.175,196 Additional laboratory studies suggest that the regulatory signal could be viscosity rather than oncotic pressure.195,197,198 Cholesterol biosynthesis also has been investigated.199,200 These studies show increased incorporation of 14C from labeled mevalonate into cholesterol by the liver in experimental nephrosis. Although this result is consistent with the interpretation that rates of hepatic cholesterol synthesis are increased in nephrosis, artifactual changes due to the addition of exogenous substrate (mevalonate) could not be ruled out in these experiments.
Lipid Transport
Several aspects of lipid transport may be impaired in nephrosis. The major cholesterol-transporting protein associated with the LDLs in the plasma is apo B-100.201 This also has been implicated as a significant apolipoprotein in atherogenesis. A recent study of nephrotic patients found that elevated serum concentrations of cholesterol, triglycerides, and phospholipids resulted mostly from changes in apo B-100-containing lipoproteins. The size of the apo B-100 pool in patients was two to three times that found in healthy subjects or in patients in remission. Fractional catabolism was decreased only slightly, suggesting that the major problem was overproduction rather than decreased breakdown.180 Hepatic uptake of LDLs may be decreased202 if the structural composition of LDLs in the circulating pool is abnormal, or if systemic neutralization of membrane negative charge leads to less efficient uptake of the largely cationic liposomes203; this would exacerbate hypercholesterolemia by decreasing negative feedback affecting hepatic synthesis. Alternatively, decreased hepatic uptake of LDLs could result from, rather than cause, hepatic overproduction of cholesterol.201 Altered transport could also occur at the level of lipoprotein receptor expression, because the nephrotic liver shows increased endocytic HDL receptor mRNA and protein but decreased expression of scavenger receptor-BI expression. The latter may be regulated by a decrease in expression of PDZ-containing kidney protein-1, which protects SR-BI from degradation.204
Metabolism of Lipids
At least one report indicates that although LDL synthesis may be increased in nephrosis, VLDL catabolism is decreased.194 Another study demonstrates that apolipoprotein E-rich IDL from nephrotic patients, but not from normal controls, inhibits sterol synthesis and cholesterol esterification.205 This finding suggests that cellular apo E metabolism may be deranged in nephrosis. Consistent with this finding, genetic variations in the expression of apo E alleles may influence the degree of lipid abnormality in nephrotic patients.206
Interest regarding catabolism of lipids in nephrosis has focused on two enzymes: LPL, which facilitates the breakdown of ester bonds in glycerides, and LCAT, which catalyzes the reaction of lecithin and cholesterol to form lysolecithin and cholesterol ester.207 In nephrotic children, elevated serum lipid levels correlate with decreased postheparin LPL activity.208 In another study of nephrotic patients, most of whom had MCNS, hepatic LPL activity was normal, but serum and adipose tissue LPL activities were decreased in association with elevated plasma triglycerides.184 Decreased hepatic209 and adipose tissue208 LPL activity in experimental rat models of nephrosis may contribute to altered lipoprotein levels in these animals. Such decreased activity may reflect decreased endothelial LPL expression.210 LCAT activity also is decreased in experimental nephrosis,211 with levels appearing to correlate with serum albumin concentration.212
Activity of these enzymes may be affected both directly and indirectly by urinary protein loss. Albumin binds to free fatty acids (FFAs); decreases in serum albumin concentration lead to FFA accumulation, thereby inhibiting LPL activity.213 LPL activity also may be inhibited by cholesterol.212 LCAT activity is inhibited by the accumulation of triglyceride and cholesterol esters,214 suggesting that abnormal LCAT activity could be a result, rather than a cause, of nephrotic
hyperlipidemia. However, lysolecithin, a reaction product that binds to albumin, inhibits LCAT activity in vitro; this feedback mechanism is blocked by the addition of physiologic levels of albumin.214,215 Therefore, the urinary loss of albumin may lead to the inhibition of lipolytic enzyme function.
hyperlipidemia. However, lysolecithin, a reaction product that binds to albumin, inhibits LCAT activity in vitro; this feedback mechanism is blocked by the addition of physiologic levels of albumin.214,215 Therefore, the urinary loss of albumin may lead to the inhibition of lipolytic enzyme function.
Albumin loss does not account entirely, however, for the elevated lipid levels. Although the infusion of albumin decreased serum lipid levels in an acute animal model of nephrotic syndrome, normalization occurred only after simultaneous infusion of heparin. This suggests the need for an additional factor that aids in clearing lipid from the plasma.216 In further experiments with this model, nephrectomy resulted in greater improvement of the hyperlipidemia than did albumin infusions alone,217 indicating that the factor may be lost in the urine. Further support for the loss of a specific regulatory molecule in the urine is provided by the observation that alteration of dietary protein intake markedly modulates the hepatic albumin synthetic rate but does not alter the hepatic synthesis of lipoproteins.218 In this study, lipoprotein synthesis correlated directly with the urinary clearance of albumin, suggesting that albumin, or another substance lost in parallel with albumin, was needed to suppress lipoprotein synthesis. Experiments with analbuminemic rats indicate that albumin itself is not likely to be the critical molecule.219 It has been suggested that the lost factor is LCAT.177 HDL, which plays an essential role in catabolism of VLDL, also may be lost in the nephrotic urine.220,221 Conversely, other studies indicate that HDL excretion is low,196 especially in MCNS.222 Apo C-II also may be lost in the urine.201,223
Clinical Significance
Regardless of the cause, the clinical significance of the lipid abnormalities in nephrosis must be considered. Hyperlipidemia has been associated with cardiovascular disease in otherwise healthy young adults, but studies evaluating such a correlation in nephrotic patients produced conflicting results. Premature coronary atherosclerosis224 and a high incidence of myocardial infarction and other cardiovascular diseases225,226 have been documented in nephrotic subjects, as well as a higher incidence of hypertension in nephrotic men than in control subjects.227 Intimal-medial thickness ratios, as an index of atherosclerotic plaque formation, were increased as a function of the number of relapses that had occurred in children and young adults with steroid-sensitive nephrotic syndrome.228 In addition, plasma levels of lipoprotein(a), a strong risk factor in cardiovascular disease, are increased.229 Macrophage morphology and function are altered by the hyperlipidemia of nephrosis230 and oxidized lipids are increased in the nephrotic syndrome,231,232 potentially contributing to the development of atheromatous plaques. In contrast, other studies have not confirmed a predisposition to atherosclerosis in patients with nephrosis.233,234 These discrepant results may reflect limitations of population base or selection bias.184,222 In the studies that did not demonstrate an increased risk, it is unclear whether stratification of the patients into cohorts according to the degree of lipid abnormality would have shown an increased risk in patients with the highest consistent elevations in lipoprotein levels. Age, underlying diagnosis, disease course, and incidence of other complicating factors such as hypertension also may be important. Another significant consideration is the possible ameliorating effect of HDL on hyperlipidemia.233,234,235,236 In several studies,184,222 HDL levels were normal or increased in MCNS. This could have a protective effect and could decrease the likelihood of cardiovascular complications. A further study of larger groups of nephrotic patients would allow differentiation among patients with other cardiac risk factors in addition to the potential hazard of elevated serum lipid levels.
A second risk involves the role of lipids in causing or enhancing the progression of the renal disease itself.237 Rats with PAN nephrosis fed high-cholesterol diets develop mesangial foam cells and mesangial proliferative changes.238 The relationship of systemic hypertension and hyperlipidemia to atherosclerosis parallels the relationship of intraglomerular hypertension and high lipid levels to focal sclerosis.238 Effective therapy of hyperlipidemia ameliorates single-nephron hyperfiltration239 and retards progression of renal failure in obese Zucker rats240 and in nephrotic rats with reduced renal mass.241 In obese Zucker rats, a relative decrease in polyunsaturated fatty acids (PUFAs), rather than high cholesterol levels, may be the most important lipidrelated factor in the progression of renal disease, because dietary supplementation with n-6 PUFA (sunflower oil) or n-3 PUFA (fish oil) slowed the progression of renal disease but only fish oil decreased serum cholesterol levels.239
In view of these considerations, and the fact that treatments for nephrosis such as steroids and diuretics may exacerbate hyperlipidemia, clinicians have invested increasing effort in controlling the lipid abnormalities of nephrosis.242 Traditional dietary therapy is of marginal value, and may actually worsen the hyperlipidemia.243 Cholestyramine may, by increasing the secretion of cholesterol into the bile, predispose one toward the development of cholesterol gallstones.201 Nicotinic acid has significant side effects and has not been studied extensively. Probucol may cause concomitant loss of HDLs.243 However, it has been shown experimentally to reverse lipid-mediated vasoconstriction244 and to be effective in treating patients who were 5 to 20 years old.245 Two other classes of drugs found to be effective in treating nephrotic hyperlipidemia are fibric acids and HMG CoA reductase inhibitors. Gemfibrozil, a fibric acid, caused a 51% reduction in serum triglyceride levels but only a 15% decrease in cholesterol when given at a dose of 600 mg twice a day to adult nephrotic patients; a 26% reduction in apo B was achieved.246 Lovastatin, an inhibitor of HMG CoA reductase, caused a 27% to 29% reduction in total cholesterol, LDL cholesterol, and apo B at a dose of 20 mg twice daily in adult patients with nephrosis due to MCN or other diseases.247 In patients with nephrotic-range
proteinuria, doses up to 40 mg twice daily caused similar decreases regardless of whether the patients were on corticosteroid therapy. A slight increase was noted in serum HDL concentrations.248 Kinetic studies showed that lovastatin enhances the catabolism of VLDL triglycerides and lowers LDL cholesterol by decreasing input rates for LDLs,249 most likely through the inhibition of LDL-apo B synthesis from VLDL.250 Atorvastatin also is effective in reducing nephrotic hyperlipidemia.251 In children, HMG CoA reductase inhibitors may be effective, but some clinicians have urged caution regarding their use in the very young child, raising the possibility that inhibiting cholesterol synthesis might impair neural myelination. This potential concern needs to be weighed against the more immediate issues of cardiovascular complications and renal disease progression.
proteinuria, doses up to 40 mg twice daily caused similar decreases regardless of whether the patients were on corticosteroid therapy. A slight increase was noted in serum HDL concentrations.248 Kinetic studies showed that lovastatin enhances the catabolism of VLDL triglycerides and lowers LDL cholesterol by decreasing input rates for LDLs,249 most likely through the inhibition of LDL-apo B synthesis from VLDL.250 Atorvastatin also is effective in reducing nephrotic hyperlipidemia.251 In children, HMG CoA reductase inhibitors may be effective, but some clinicians have urged caution regarding their use in the very young child, raising the possibility that inhibiting cholesterol synthesis might impair neural myelination. This potential concern needs to be weighed against the more immediate issues of cardiovascular complications and renal disease progression.
Disorders of Hemostasis
The association between nephrotic syndrome and intravascular coagulation has been known for more than a century, but it was not until 1948 that the concept of a thrombotic diathesis in nephrotic patients was proposed.252 In a review of 3,377 children with nephrotic syndrome, the incidence of thromboembolic complications was 1.8%.253 The prevalence of such complications in adult nephrotic subjects is much higher and averaged 26% in eight series of patients.254 Thrombosis may occur at any stage during the course of the nephrotic syndrome, but it is most frequent in the early months.
Extent of Clinical Involvement
Deep vein thrombosis of the leg is the most common thrombotic complication in the nephrotic adult and was responsible for one third of the thromboembolic complications in the largest published series of nephrotic children.253 Other reported sites of venous thromboses include the subclavian, axillary, external jugular, portal, splenic, hepatic, and mesenteric veins as well as the superficial cerebral cortical sinus, where thrombosis has been observed in both children and adults and may be fatal.255,256 Arterial thrombosis occurs less frequently and is seen primarily in children. Thrombosis of the aorta and of the mesenteric, axillary, femoral, ophthalmic, carotid, cerebral, renal, pulmonary, and coronary arteries has been reported, as has intracardiac thrombosis.255,256 The pulmonary257 and femoral arteries are particularly susceptible, the former potentially resulting in infarction258 and the latter usually as a complication of attempted blood sampling from the femoral vein. Although recanalization of the artery does occur, a relatively high proportion of patients with arterial thrombi die.257
The lesion that has attracted the most attention, however, is RVT. It is most often seen with membranous glomerulopathy259 to the extent that at one time there was some controversy about whether RVT was the cause, rather than a complication, of the glomerular lesion. The reported frequency of RVT in patients with membranous disease has ranged from 4% to 51%, depending on the methods used to establish the diagnosis and to select the patient population for study.260 The mean prevalence is 12%. There is a high incidence of RVT in mesangiocapillary glomerulonephritis and the nephritis of systemic lupus erythematosus, and RVT can complicate numerous other renal diseases.255 It is relatively uncommon in nephrotic children except in those with congenital nephrotic syndrome of the Finnish type.261
The thrombosis may involve only the renal venous system or it may extend into the inferior vena cava. Therefore, it is not surprising that pulmonary emboli develop in about 40% of adult patients with RVT, although pulmonary emboli rarely occur in children. Death from pulmonary emboli is uncommon.253,255
The diagnosis of acute RVT is suggested by flank pain, costovertebral angle tenderness, gross hematuria, increased proteinuria, and acute reduction in renal function; intravenous pyelography may show ureteral notching or pelvicaliceal irregularities.260 Ultrasonography may demonstrate only a large kidney or may visualize the thrombus if it extends into the renal vein or inferior vena cava. However, Doppler ultrasound analysis often shows decreased venous blood flow. A more chronic form of RVT may be asymptomatic and may be identifiable only by venography.260,262 The mode of presentation of other thromboses depends on their site. Diagnosis can be difficult and the existence of arterial thrombosis may not be realized until autopsy. Ultrasonography and angiography are the preferred studies.
Regulators of Coagulation
The blood coagulation pathway represents a cascade of events that regulate the dynamic balance between the ability of the blood to remain fluid and its tendency to assume a gelled state in the presence of altered flow conditions or exposure to nonendothelial surfaces. Contributing to hemostatic balance are several opposing systems that contribute to a cascade through which a series of enzymes regulates fibrin polymerization (Fig. 52.6). Coagulation is initiated by the activation of prekallikrein to kallikrein (intrinsic pathway), or by exposure to nonendothelial tissues (extrinsic pathway). These pathways meet in the activation of factor IX, initiating a common pathway in which a central role is played by thrombin (factor IIa). This enzyme stimulates the activation of fibrinogen to fibrin and the aggregation of platelets. It also activates factors V, VIII, and XIII. Factor XIIIa triggers the cross-linking of fibrin monomer into a stable polymer. Two systems oppose clot formation and stability. Protein C is processed to activated protein C (aPC) by thrombin complexed with thrombomodulin. With free protein S as a cofactor, aPC inactivates factors Va and VIIIa. The other system that opposes coagulation is the fibrinolytic pathway, in which plasminogen activators convert plasminogen to plasmin, which degrades fibrin polymer. Several proteins inhibit these pathways: antithrombin III and α2-macroglobulin inhibit
thrombin; α1-antiplasmin and α2-macroglobulin inhibit plasmin; and the plasminogen activator inhibitors (PAIs) inhibit plasminogen activators and aPC. Activated protein C, in turn, opposes PAI effects. Many of the components of these pathways are altered in nephrosis. In addition, physical conditions of the nephrotic syndrome, such as venous stasis, hemoconcentration, increased blood viscosity, and possibly the administration of steroids, may also contribute to enhanced blood clotting. These nephrotic effects on coagulation pathways, which are listed in Table 52.3 and discussed in detail elsewhere,255,262,263 are considered here briefly.
thrombin; α1-antiplasmin and α2-macroglobulin inhibit plasmin; and the plasminogen activator inhibitors (PAIs) inhibit plasminogen activators and aPC. Activated protein C, in turn, opposes PAI effects. Many of the components of these pathways are altered in nephrosis. In addition, physical conditions of the nephrotic syndrome, such as venous stasis, hemoconcentration, increased blood viscosity, and possibly the administration of steroids, may also contribute to enhanced blood clotting. These nephrotic effects on coagulation pathways, which are listed in Table 52.3 and discussed in detail elsewhere,255,262,263 are considered here briefly.
 FIGURE 52.6 Interactions among circulating participants in the coagulation cascade. The intrinsic pathway begins in the upper left; the contact system of factor XII, prekallikrein, and high-molecular-weight kininogen (HMWK) that initiates the intrinsic cascade is shown in condensed form. The extrinsic pathway begins in the upper right. The action of factor VIIa on the activation of factor IX has blurred the distinction between these two limbs of the cascade. The common pathway begins with activation of factor × and results in thrombin activity and fibrin cross-linking. Coagulation is counteracted by the fibrinolytic pathway, including plasminogen activators and plasminogen/plasmin. In addition, there are several anticoagulant pathways, most notably inhibition of factors Va and Villa by the inhibitory cofactors protein S and activated protein C (aPC). Alpha-2-macroglobulin, which binds to and inhibits most of the enzymes in this system, is not shown here for the purpose of simplicity. The relationship between heparin sulfate proteoglycans (HSPGs) and plasminogen activator activity shown in the lower right should be regarded as hypothetical for human pathophysiology. Thick solid lines with arrows indicate reactions; thin solid lines with triangular arrows denote catalytic effects; joining lines show cofactors in catalysis. Broken lines with fork-tailed arrows represent inhibitory actions. PAI-1 and PAI-3 bind to aPC; this interaction may cause mutual inhibition of the action of these proteins. α1 AP, α1-antiplasmin; ATIII, antithrombin III; C4bBP, complement factor 4b-binding protein; PAIs, plasminogen activator inhibitors; TM, thrombomodulin; tPA, tissue-type plasminogen activator; uPA, urokinase plasminogen activator. (Figure composed with the assistance of G.A.Soff.) |
TABLE 52.3 Coagulation System Abnormalities in the Nephrotic Syndrome | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Platelet Aggregation
Platelets may play a role in the genesis of the coagulopathy of nephrotic syndrome. Thrombocytosis is commonly found, especially early in the disease course,264 and platelets show markers of activation.265 Platelet aggregability is increased and platelet degranulation has been described.266 In addition, plasma levels of the platelet release substance β-thromboglobulin are increased.264,267 Levels of platelet factor 4 are normal268 or increased.264 In contrast, platelet calcium ion release and ATP secretion have been found to be decreased in nephrosis.269 The authors of that report suggest that platelets may become desensitized to platelet activating factor (PAF) because of exposure to consistently high ambient concentrations.
Platelet hyperaggregability correlates with the degree of proteinuria and with plasma cholesterol levels. It can be reversed by the addition of urine protein.268 These findings suggest that the urinary loss of albumin270 or of some factor that normally inhibits platelet aggregation is responsible for the changes seen in nephrotic syndrome. Alternatively, hyperlipidemia could result in the changes, because platelet aggregation is increased in patients with type II hyperlipoproteinemia to a degree that is comparable to that seen in nephrotic syndrome.271 Altered platelet function could be a response to hypoalbuminemia, because the conversion of arachidonic acid into metabolites that aggregate platelets is known to be regulated by albumin.272,273 Thus, platelets show greater production of thromboxane B2 and malondialdehyde in nephrotic plasma than in normal plasma when challenged with arachidonic acid. The addition of albumin to the nephrotic plasma corrects this abnormality.274 Finally, it is possible that alterations in platelet membranes could be responsible for increased platelet activity. Platelet membranes contain a sialoglycoprotein with a pK of 1.8 to 2.2.275 This may be important in preventing spontaneous platelet aggregation or platelet interaction with the vessel wall.276 Because systemic negative-charge sites may be reduced during relapses of nephrotic syndrome,69 the same mechanism responsible for the reduction of negative-charge sites in the GBM could enhance platelet aggregation.
Coagulation Factors
Evidence that various functions of blood coagulation are activated in nephrosis is provided by increased concentration of the D-dimer of fibrinogen.277 Elevated levels of this breakdown product of cross-linked fibrin indicate that both the coagulation and fibrinolytic pathways are concurrently activated. Plasma fibrinogen is consistently elevated in nephrotic syndrome due to increased hepatic synthesis. Chromatography demonstrates both increased polymerization and increased proteolytic derivatives of fibrinogen or fibrin. These changes reverse as patients with nephrotic syndrome enter remission.278 This evidence for increased intravascular fibrin formation is supported by the finding of increased plasma levels of fibrinopeptide A, at least in FSGS.278
The concentration in nephrotic patients of coagulation factors that initiate fibrin formation likely reflects the balance between the increased hepatic synthesis of these proteins, triggered as part of a nonspecific response to hypoproteinemia, as described previously for lipoproteins, and urinary losses. Therefore, lower molecular-size proteins (approximately less than 70 kDa) may be lost in the urine, whereas higher molecular-size proteins (greater than 300 kDa) are likely to be increased in the plasma. For example, most studies agree that levels of factors V and VIII are increased in the plasma, whereas those of factors IX, XI, and XII are decreased279 despite the possibility that production may still be increased. The magnitude of the increase in concentration of factors V and VIII, for example, correlates with the degree of reduction in serum albumin and the likely resulting increased hepatic synthesis of these factors, stimulated by hypoalbuminemia.280 Plasma levels of factors II, VII, X, and XIII are often found to be increased.281,282,283 There is no direct evidence that any of these changes are responsible for the hypercoagulable state. Indeed, the alterations in blood levels of these factors are often inconsistent and of minor degree. Therefore, these abnormalities may be of more biochemical than clinical interest. Most of the changes in concentration of these zymogen factors reverse with clinical remission of the nephrotic syndrome.
Inhibitors of Coagulation
The most well-studied biologic antagonist of coagulation, antithrombin III, is decreased in the plasma of nephrotic patients.284,285,286,287,288 This is presumed to be due to urinary loss of antithrombin III, which has a relatively low molecular weight. Indeed, plasma antithrombin III levels in nephrotic syndrome correlate well with those of serum albumin and inversely with the renal clearance of antithrombin III. Because hereditary antithrombin III deficiency is associated with frequent thrombosis, it was hypothesized that the low plasma antithrombin III levels were insufficient to inactivate procoagulant factors and were the major cause for the hypercoagulable state and the development of thrombosis in nephrotic syndrome.286 However, only patients with plasma albumin levels below 2 g per deciliter show significant reductions in plasma antithrombin III levels,289 whereas hypercoagulability may be present in patients with albumin levels exceeding this value. Further, normal plasma levels of antithrombin III were found in nephrotic subjects who had loss of antithrombin III in the urine and who had thromboembolic complications.290 Indeed, decreases in antithrombin III levels may be compensated for by increased plasma levels of α2-macroglobulin,283 leading to increased total antithrombin activity.278
A complex effect of nephrosis has been noted on the anticoagulation pathway by which protein C is activated by thrombin and thrombomodulin to aPC. Activated protein C and protein S combine to inhibit factor VIIIa (decreasing activation of factor X) and factor Va (decreasing activation of factor XI). Protein S exists in circulation in two forms: free and bound to C4b binding protein. Only the free protein S can serve as a cofactor with protein C to inactivate factors Va and VIIIa.291 In nephrotic syndrome, although small amounts of protein C are lost in the urine, serum concentrations are normal,292,293 indicating that hepatic synthesis is able to compensate. In contrast, although plasma levels of C4b binding protein-protein S complex are usually normal to elevated, free levels are markedly decreased.294 This finding is consistent with the molecular sizes of the complex (640 kDa) and the free protein S (69 kDa). Acquired dysfunction of this system is a common cause of thrombotic diathesis.291,295 Indeed, intractable deep vein thrombosis in an unusual nephrotic child with decreases in both protein S and C4b-binding protein concentrations296 supports the notion that abnormalities of this system are clinically significant.
Fibrinolysis
Alterations in the concentrations of several of the components of the fibrinolytic system have been documented.297 Decreased fibrinolytic activity has been associated with hypertriglyceridemia.298 Of the individual components of the fibrinolytic system, decreased concentrations of plasminogen have been found287,299; levels of tissue-type plasminogen activator277 and PAI-1300 are elevated. Varying levels of α2-antiplasmin have been reported,277,278 possibly depending on whether thrombosis has occurred. Of the serine protease inhibitors that modulate both the fibrinolytic and thrombin systems, levels of α2-macroglobulin are increased and those of α1-antitrypsin are decreased.283 Again, this probably reflects the effect of urinary loss on plasma concentrations. It also is possible that local vascular conditions affect fibrinolysis. The infusion of a variety of polyanions causes an immediate local increase in the release of plasminogen activator and PAI activity in the sow ear. The PAI activity immediately returns to normal but the increase in plasminogen activator activity is sustained,301 suggesting that negative charges in the vascular tree are important for inducing the plasminogen activator pathway. If nephrosis is associated with a generalized reduction of fixed negative charge sites in the vascular space,69 it is likely that decreased negative charges could have an impact on the induction of fibrinolysis. Coupled with increased α2-antiplasmin concentrations, decreased tissue-type plasminogen activator activity would significantly impair fibrin degradation. However, the increase in circulating D-dimer cited previously indicates that at least some fibrinolysis occurs in nephrosis. It also is possible that fibrin polymerization is impaired in nephrotic patients. In one study of a broad spectrum of adults with nephrosis, most had prolonged thrombin times. Half of the patients showed decreased ability to polymerize fibrin monomer. There was no correlation of this finding with prothrombin time, partial thromboplastin time, fibrin degradation products, antithrombin III concentration, or platelet count.302
Physical Factors Affecting Coagulation
Physical factors such as increased blood viscosity also may contribute to the generation of thromboembolic complications.303 Both children and adults with MCNS and well-preserved renal function may have marked hemoconcentration with elevated hematocrit and hemoglobin concentrations. Such changes are associated with disproportionate increases in viscosity and could be aggravated by the therapeutic use of diuretics, especially if these cause further hemoconcentration. In addition, when plasma fibrinogen levels increase, especially to values as high as 1 g per deciliter as can be seen in nephrotic syndrome, they cause increased erythrocyte aggregation and marked increases in plasma viscosity.304 A role for physical factors is supported by the high incidence of renal vein thrombosis, because hemoconcentration of the blood and the effect of urinary inhibitor loss will be most pronounced in the radicles of the renal vein.305
Steroid administration increases the concentrations of several clotting factors and modifies coagulation mechanisms.255 Moreover, a high incidence of thromboses was recorded after these drugs were first used to treat nephrotic syndrome.306 Both arterial and venous thromboses, however, have been found in nephrotic subjects not receiving steroids. Furthermore, a hypercoagulable state is present in untreated MCNS patients, and levels of the coagulation factors do not change after steroid treatment is implemented.278
Except for the protein S data, the potential relationship between various biochemical findings and the clinical importance of thrombus formation remains largely theoretical. For example, the biochemical abnormalities in children may be more severe than in adults with the nephrotic syndrome, whereas the incidence of thromboembolic phenomena is worse in adults.307 This may reflect the fact that MCN is more common in children, whereas membranous nephropathy (see Chapter 63) is more common in adults. Therefore, the underlying nature of the disease may be important in determining the occurrence of intravascular coagulation. Finally, it is important to consider the physiologic conditions within the circulatory tree. For example, one study suggested that although platelet aggregation is increased in nephrosis, fibrin conversion inhibits platelet interaction with the vessel wall extracellular matrix, actually decreasing the likelihood of platelet participation in thrombus formation. In this system, the data suggest that increased fibrin formation, but not increased platelet aggregation, contributes to the hypercoagulability of the nephrotic syndrome.308 In support of a primary role for the coagulation cascade, nephrotic patients show biochemical evidence for endothelial injury.309 The effects of negative-charge sites must also be considered. Like plasminogen activator activity, antithrombin III is active in association with vascular wall heparin sulfate proteoglycans. If MCNS involves a generalized decrease in negative-charge sites, antithrombin III activity may be impaired.
Infections
It has long been known that patients with MCNS have increased susceptibility to infection. This increase may be related to the prolonged presence of gross edema or ascites, which are composed of fluids that represent ideal culture media for bacterial growth. The infection risk may be potentiated by therapy with steroids or immunosuppressive drugs, although the high incidence of infections was noted in the era before these drugs were available. Humoral responses to bacteria may be defective. Plasma concentrations of IgG are markedly reduced during relapse,310 and the ability of MCN patients to generate specific antibodies is impaired311 between as well as during relapses. Although the role of these abnormalities in predisposing nephrotic subjects to infections remains to be elucidated, it may be significant that boys with MCN respond poorly to the hepatitis B vaccine.312 This same population has a higher incidence of chronic hepatitis B surface antigenemia than that found in a control population.313 Another factor that could contribute to a high rate of infections in patients with nephrosis is a decreased serum level of alternative complement pathway factor B. Absence of this factor has been linked to defective opsonization of Escherichia coli in nephrotic patients314 and to defective neutrophil function.315 Serum levels of hemolytic factor D also are decreased in patients during relapse and return to normal with remission.316 Levels of both of these factors correlate strongly with serum albumin concentration, suggesting that decreased serum levels result from urinary loss. These concerns have led to recommendations that both pediatric and adult patients with nephrosis should receive the pneumococcal polysaccharide (23-valent) vaccine.317,318 In addition, the use of penicillin prophylaxis may be required in children younger than 2 years of age, or in older patients who have low antibody titers or recurrent pneumococcal infections.319 Immune system abnormalities more specifically associated with MCN are unlikely to result entirely from albuminuria, because they are specific for that disease; these will be considered in the section on MCN later in this chapter.
Consequences of Loss of Other Proteins
Numerous proteins in addition to albumin are lost in the urine. In most instances, these proteins are of a similar or smaller size than albumin. Such losses could alter function in the endocrine system or in metabolic pathways. Therefore, the loss of insulinlike growth factors could contribute to poor growth in some nephrotic children.320 The urinary loss of thyroxin-binding globulin (TBG) correlates well with total urinary protein excretion.321 In addition to TBG, losses of thyroxine (T4) and triiodothyronine (T3) in nephrotic urine are associated with decreased serum levels of T3 and TBG. Most of the patients studied were clinically euthyroid and their serum levels of free T4 and thyroid-stimulating hormone (TSH) did not differ from those in normal control subjects.322 In addition, their values for T3 uptake were normal. Another study documented urinary losses, but normal serum concentrations of TBG.323 The patients had low or lownormal T4 levels. Such differences in findings could relate to the underlying cause for nephrotic syndrome, whether it is associated with selective or nonselective proteinuria and whether it is accompanied by uremia. Children with MCNS have serum T4 or free T4 levels that are marginally low.324 They have been interpreted as having mild thyroid failure based on increased baseline TSH levels and their response to thyrotropin-releasing hormone.325 In patients who have congenital nephrotic syndrome, nephrectomy to eliminate proteinuria is associated with normalization of thyroid status, indicating that these abnormalities result from massive proteinuria rather than an intrinsic glandular defect.326
Total serum calcium is markedly reduced, primarily because of hypoalbuminemia and the consequent decrease in protein-bound calcium. Serum ionized calcium levels may be reduced as well,327,328 even in nephrotic subjects with normal renal function; this may result in symptomatic hypocalcemia. At least some of the reduction in ionized calcium is the consequence of a loss of 25-hydroxyvitamin D (25-[OH]-D) in nephrotic urine329,330; normally bound to and absorbed via megalin, vitamin D-binding protein is lost as the protein reabsorptive capacity is saturated and exceeded due to proteinuria. Other metabolites of vitamin D may be lost as well.331 Low plasma levels of 25-(OH)-D, 1,25(OH)2-D, and 24,25-(OH)2-D have been reported in patients with nephrotic syndrome,330 and intestinal absorption of calcium is reduced327; serum
parathyroid hormone levels are increased.328 From these observations, it has been postulated332 that urinary loss of vitamin D complex results in decreased absorption of intestinal calcium, skeletal resistance to parathyroid hormone, and reduced serum calcium levels. In turn, these changes cause increased parathyroid hormone (PTH) production and could result in defective bone mineralization. Although it has not been proved that the changes described cause significant bone disease,333 ongoing steroid therapy may increase the likelihood of clinically significant problems.334,335 In 60 children and adolescents with relapsing nephrotic syndrome, bone mineral density (BMD) was decreased, but whole-body bone mineral content was higher because of an increase in the body-metabolic index.335 In patients with unremitting proteinuria and low 25-OH-D levels, supplementation with oral calcium and 25-OH-D should be considered.336
parathyroid hormone levels are increased.328 From these observations, it has been postulated332 that urinary loss of vitamin D complex results in decreased absorption of intestinal calcium, skeletal resistance to parathyroid hormone, and reduced serum calcium levels. In turn, these changes cause increased parathyroid hormone (PTH) production and could result in defective bone mineralization. Although it has not been proved that the changes described cause significant bone disease,333 ongoing steroid therapy may increase the likelihood of clinically significant problems.334,335 In 60 children and adolescents with relapsing nephrotic syndrome, bone mineral density (BMD) was decreased, but whole-body bone mineral content was higher because of an increase in the body-metabolic index.335 In patients with unremitting proteinuria and low 25-OH-D levels, supplementation with oral calcium and 25-OH-D should be considered.336
Carbohydrate metabolism also may be deranged. Of 38 adult nephrotic patients who had not received any drugs, including glucocorticoids, for at least 2 months, 14 had oral glucose tolerance test results that were similar to those found in diabetic patients.337 Affected patients had increased insulin secretion that was thought to be secondary to increased growth hormone levels. The initiating event for these changes was not determined. There was no correlation of these findings with either serum albumin levels or renal histopathology. This observation raises the question of whether nephrotic hyperlipidemia could contribute to the development of noninsulin-dependent (type 2) diabetes mellitus.
Alterations in trace metal metabolism may be due to urinary losses of either the metals or their carrier proteins. Decreased serum levels of both iron and copper, associated with a low serum iron-binding capacity and low erythrocyte copper content, have been documented in nephrotic syndrome.338 Serum levels of copper, but not iron, were improved by oral administration of the metal. Urinary iron and copper concentrations correlated with protein excretion. The intravenous infusion of albumin led to increased albuminuria and increased metal excretion. In each case, the abnormalities appeared to be related to urinary protein loss. Nephrotic children may develop anemia secondary to urinary loss of transferrin and iron.339,340 Serum zinc levels are low in nephrotic syndrome, but urinary excretion of zinc is not elevated. Zinc binds to albumin so that serum zinc levels change with alterations in albumin levels, regardless of the etiology of the nephrotic syndrome. However, decreased zinc content in the hair of these patients suggests that other aspects of zinc metabolism also may be deranged.341
Many drugs are protein bound in the plasma. Hypoalbuminemia decreases the number of drug-binding sites and could result in increased toxicity of drugs that normally are bound to protein. For example, digoxin is 25% bound to proteins in the plasma, digitoxin is 90% bound, hydrochlorothiazide is 60% bound, and furosemide is 96% bound; hydralazine, prazosin, and diazoxide are all approximately 90% bound, whereas the binding of barbiturates varies from 5% to 80% depending on the molecular structure.342
General Approach to the Treatment of Nephrotic Syndrome
Based on the consequences of proteinuria described in the preceding text, symptomatic treatment regimens have been developed for the care of all nephrotic patients, even beyond those classified as having a form of primary nephrotic syndrome. These treatments, which are aimed at the physiology of nephrosis rather than the etiology of the disease, have implications beyond the reduction of nephrotic proteinuria and edema.
Diuretics and Fluid Management
Many patients with nephrotic syndrome respond to the acute use of diuretics with increased urinary losses of sodium and water. Although diuretics may be effective, there are limited indications for their use in the acute treatment of nephrotic subjects, especially children. The degree of diuresis and natriuresis they induce is small compared to that observed when the patient responds to treatment directed at the underlying cause. Furthermore, it is possible that diuretic use, by depleting intravascular as well as interstitial fluid volume, may contribute to the development of shock seen in some patients with MCNS.173
Nonetheless, oral or parenteral diuretics are effective and often are indicated in the management of persistent edema. Parenteral furosemide is more effective than an orally administered drug. Treatment may be initiated at 1 mg per kilogram (up to 40 mg) with judicious increases up to three to four times the usual dose343 in an effort to elicit a response. Diuretic therapy may be less effective in patients with primary nephrotic syndrome than in most other patients, in part because of a combination of factors resulting in a physiologically decreased ability to excrete sodium. This is especially true of the loop diuretics such as furosemide, which also may be inhibited by its binding to albumin present in the tubular lumen.344 Although spironolactone interferes with distal nephron sodium reabsorption and thus is a theoretically useful diuretic, in practice its delayed onset of action and relatively weak potency limit its usefulness to being a potentiating agent with loop diuretics. Metolazone, a thiazidelike drug that impedes both proximal and late distal nephron sodium absorption,345 is a singularly effective oral diuretic in patients with sodium retention secondary to nephrotic syndrome. There is a possibility that diuretic therapy will deplete the intravascular volume without significantly reducing the tissue edema. Therefore, diuretics should be administered with care and withheld in patients for whom a rapid response to steroids is anticipated.
If the patient has anasarca, if respiratory embarrassment results from ascites or pleural effusion, if scrotal or vulval edema is sufficiently severe to threaten tissue breakdown,346 or if peritonitis is present, then more aggressive therapy is warranted to decrease the amount of edema. A useful regimen consists of oral spironolactone, 1 mg per kilogram per day, and daily intravenous infusions of albumin, 0.5 g per
kilogram initially and increasing, if well tolerated, to 1 g per kilogram per day. The albumin infusion should be preceded by the intravenous infusion of furosemide, 0.5 mg per kilogram. A repeated dose of diuretic is given toward the end of the albumin infusion. Blood pressure should be monitored throughout the albumin infusion to help avoid complications from rapid mobilization of edema fluid into the circulation, although the regimen usually is free from significant side effects when used in children and young adults. Some observations suggest that the administration of albumin may result in more severe glomerular epithelial changes, should raise the oncotic pressure of the tissue space, should delay the response to corticosteroid therapy, and should induce more frequent relapses after remission.347 In view of the potential for complications of albumin infusion,348 this treatment should be reserved for the specific indications of respiratory embarrassment, tissue breakdown, or the need to elicit urine output to confirm the diagnosis of nephrosis.
kilogram initially and increasing, if well tolerated, to 1 g per kilogram per day. The albumin infusion should be preceded by the intravenous infusion of furosemide, 0.5 mg per kilogram. A repeated dose of diuretic is given toward the end of the albumin infusion. Blood pressure should be monitored throughout the albumin infusion to help avoid complications from rapid mobilization of edema fluid into the circulation, although the regimen usually is free from significant side effects when used in children and young adults. Some observations suggest that the administration of albumin may result in more severe glomerular epithelial changes, should raise the oncotic pressure of the tissue space, should delay the response to corticosteroid therapy, and should induce more frequent relapses after remission.347 In view of the potential for complications of albumin infusion,348 this treatment should be reserved for the specific indications of respiratory embarrassment, tissue breakdown, or the need to elicit urine output to confirm the diagnosis of nephrosis.
Management of the acute phase of nephrotic syndrome should include dietary sodium restriction. During relapse, dietary sodium intake optimally should be reduced to about 0.5 g per day, which is approximately equivalent to a 1-g salt diet or about 20 mEq of sodium per day. Such severe dietary restriction is difficult to accomplish even in a carefully controlled hospital setting. It is important to emphasize that severe restriction of sodium intake will not result in weight loss when nephrotic patients are in the sodium-retaining phase of their disease. In such patients, the normal extrarenal losses of sodium may amount to less than 10 mEq per day. Therefore, severe dietary sodium restriction is intended to prevent further accumulation of edema. Use of a salt substitute may facilitate compliance with the sodium-restricted diet, but in patients with renal insufficiency, it must be limited because these preparations consist of potassium and ammonium salts.
At home, most patients can rarely manage dietary restriction below that of a no-added-salt diet. This provides a sodium intake of 40 to 60 mEq per day depending on the patient’s size. Even in remission, it should be employed not only to lessen the risk of edema formation if the patient has a relapse, but also to reduce side effects from steroid administration.
Although there is some debate regarding fluid management, we believe that fluid intake also should be restricted, at least initially. If intake equals insensible fluid losses plus urine output, the patient’s weight will remain stable without further accumulation of edema. To accomplish a loss of weight, fluid intake must be reduced below this level. Some nephrotic patients experience intense thirst. If sodium intake is limited and fluid intake is great, the patient can become hyponatremic and will remain edematous.
Anecdotal experience suggests that bed rest may potentiate a diuresis, perhaps by redistributing fluid from the peripheral tissues to the vascular space, thereby increasing renal blood flow. Bed rest also may accelerate a response to steroids and, when practical, should be advised for patients with anasarca. Other therapies that may facilitate a diuresis in some patients by mobilizing tissue fluid include local pressure using surgical elastic stockings or immersion up to the neck in warm water.135 After remission is induced, a high-protein diet may increase the rate at which plasma protein concentration returns to normal.349
Dietary Treatment
As already indicated, dietary sodium and water restriction is important in the management of the acute phase of nephrotic syndrome. Long-term reduction of dietary sodium intake in combination with diuretics can be most effective at controlling edema in patients resistant to steroids and other pharmacologic agents.
Other dietary manipulations have received attention. High-protein diets can be beneficial in special groups of patients such as those with congenital nephrotic syndrome.350 The concept of increasing dietary protein in all nephrotic subjects has not proved to be beneficial. Although albumin synthesis increases with such diets, urinary protein excretion increases too, possibly due to an angiotensin II-induced increase in glomerular permeability. Therefore, it has been proposed that inhibitors of ACE should be used in conjunction with the high-protein intake.100
Conversely, low-protein diets will decrease albuminuria and will increase albumin mass.351 This reflects conservation of essential amino acids in response to proteinuria.352 The use of soy-based, low-protein, low-fat diets rich in polyunsaturated fats and supplemented with essential amino acids or keto analogs results in decreases in urinary protein excretion and in serum total and LDL cholesterol levels.353,354,355 Supplementing diets with fish oils containing ω-3 fatty acids has not proved beneficial.356 Concerns about such dietary manipulations include their cost, the lack of patient acceptance, and whether strict adherence might result in specific nutritional deficiencies.351,357 Reducing dietary fat intake in hyperlipidemic but otherwise healthy children to the levels recommended by the National Cholesterol Education Program is safe and does not affect the children’s growth or development.358 The safety of more stringent restrictions in children with renal disease remains to be determined.
Lowering dietary fat intake has a limited effect on reducing serum lipids in the nephrotic patient. Therefore, attention has focused on whether there is a role for lipid-lowering drugs in managing hyperlipidemia in these patients (see the following paragraph). The efficacy of oligoantigenic diets or those that eliminate specific antigens has been tested in nephrotic subjects (see section on Atopy and Minimal Change Nephropathy). This approach is based on the possibility that some cases of nephrotic syndrome may be the consequence of food allergies, especially to milk and dairy products.
Lipid-Lowering Drugs
Long-term administration of HMG CoA reductase inhibitors in nephrotic patients will induce reductions of serum triglycerides as well as serum total and LDL cholesterol levels; HDL
cholesterol values are maintained.201,251,359,360 Long-term benefits from this therapy have yet to be proved, although anecdotal experience suggests that it may help to reduce proteinuria and maintain GFR,361 as observed in experimental animals (see section on Hyperlipidemia).
cholesterol values are maintained.201,251,359,360 Long-term benefits from this therapy have yet to be proved, although anecdotal experience suggests that it may help to reduce proteinuria and maintain GFR,361 as observed in experimental animals (see section on Hyperlipidemia).
Other Medical Therapy
Anticoagulation has been employed as indicated in patients with intravascular thromboses. Some clinicians use small doses of aspirin in an effort to prevent repeated thromboembolic episodes,362 but there are no studies confirming the efficacy of this approach.
Although hypocalcemia is common in relapses of nephrotic syndrome, it is likely that most patients in relapse of brief duration do not require routine treatment with calcium or with vitamin D or one of its metabolites.363 Supplementation with oral calcium (1 g per day) and 25-OH-vitamin D (25 µg per day) maintains normal bone status in steroid-dependent nephrotic patients.336 Vitamin D treatment may be more routinely necessary in patients who are steroidresistant364 (see section on Consequences of Loss of Other Proteins previously in this chapter). Similarly, patients who develop other deficiencies secondary to renal losses, such as iron-deficiency anemia, will require appropriate treatment.
Medical Treatment to Reduce Proteinuria
In patients who prove to be unresponsive to therapy directed at the underlying cause of proteinuria, efforts have been employed to decrease protein loss by employing drugs that appear to be directed against the physiology of glomerular proteinuria. The intent behind this treatment is to facilitate general medical management by decreasing proteinuria, increasing serum albumin, and thereby lessening edema formation. Thus, indomethacin is effective at ameliorating intractable nephrosis.365 Because of the association of nonsteroidal agents with renal failure in some nephrotic patients as described in the section on Complications, more recently, clinicians have emphasized the use of ACE inhibition.103,366 Each of these treatments appears to reduce glomerular capillary hydraulic pressure, by different mechanisms. ACE inhibitors act by decreasing postglomerular arteriolar resistance, whereas nonsteroidal drugs enhance preglomerular capillary resistance.367 Consistent with these mechanisms, Garini and colleagues368 compared the effects of indomethacin, captopril, and the calcium-channel blocker, nifedipine, on the changes in renal hemodynamics and proteinuria induced by a high-protein mean. The increases in GFR and renal blood flow were not blocked by captopril or nifedipine, but were blocked by indomethacin. The increase in protein excretion was blocked by indomethacin and captopril but not by nifedipine.368 Indeed, dihydropyridine calcium channel blockers do not decrease proteinuria369 and may even increase it. Therefore, the effects on hemodynamics were different and the effect on proteinuria could be distinguished from both systemic effects on blood pressure and effects on GFR and RBF. In patients with membranous nephropathy, ACE inhibition may have a size-selective effect,370 decreasing the fractional clearance of larger (>60 Å) molecules, suggesting an effect on the shunt pathway of macromolecular clearance. Such studies have supported efforts by clinicians to utilize ACE inhibitors to decrease proteinuria not only for symptomatic management but also in an effort to delay or prevent the progression of chronic kidney disease. Angiotensin-receptor blockade (ARB) has an equivalent effect to that of ACE inhibition, and the combination of ACE inhibition and ARB has been proposed to be more effective than either alone.371 The efficacy of this treatment will be considered further in the section on FSGS.
It is possible that at least a part of the antiproteinuric effect of ACE inhibition is not mediated by the glomerulus. Angiotensin blockade enhances megalin expression and albumin reabsorption,372 and indeed, the angiotensin antagonist, eplerenone, is synergistic with ACE inhibition in reducing nephrotic proteinuria.373
Nephrectomy. Unilateral nephrectomy has proved beneficial in some infants presenting with nephrotic syndrome in the first year of life,374 and bilateral nephrectomy may be a useful part of the aggressive approach required in patients with congenital nephrotic syndrome if they are to survive to an age when transplantation is feasible375,376 (see section on Nephrotic Syndrome in the First Year of Life, later in this chapter). Bilateral embolization of renal arteries can be an important therapeutic option in carefully selected patients with nephrotic syndrome who appear destined to progress to end-stage renal failure.377
MINIMAL CHANGE NEPHROPATHY
Although MCN is often thought of as a pediatric disease, it is the third most common cause of nephrosis in adults9 and remains the most common cause of nephrotic syndrome in children younger than 16 years of age. Indeed, it is the second most common primary renal parenchymal disease in that age group. Two to 7 new cases occur annually per 100,000 children,378 and the prevalence is about 15 cases per 100,000 children. Although most children with MCN achieve permanent remission of symptoms by the time they reach puberty, some cases persist into adulthood.379 Furthermore, new cases have been reported in the eighth decade of life.380 However, the relative incidence of MCN as the etiology of nephrotic syndrome decreases with age in both children and adults.9,380,381,382 Although it is not clear that adult-onset disease represents the same entity as that found in childhood, or that all patients with the clinical picture of MCNS have an identical disease, the clinical course and outcome of pediatric and adult cases appear to be sufficiently similar380 to consider all cases together.
Minimal change disease can appear in the first year of life, but it is more common later, with a peak incidence at 2 years of age. Most pediatric surveys report that it occurs
twice as often in boys as in girls, whereas it has an equal sex incidence in adults.383 Although no precipitating cause may be apparent in many children, it is not unusual for the development of edema and proteinuria to be preceded by an upper respiratory tract infection, an allergic reaction to an insect sting or other immunogenic stimuli, or the use of certain drugs (Table 52.4).383,384,385,386,387,388,398,400,401,402,403,404,405,406,407,409,410,411,415 In both adult and pediatric patients, malignancies, especially Hodgkin disease, have been associated with the development of MCNS (see section on Nephrotic Syndrome and Malignancies).
twice as often in boys as in girls, whereas it has an equal sex incidence in adults.383 Although no precipitating cause may be apparent in many children, it is not unusual for the development of edema and proteinuria to be preceded by an upper respiratory tract infection, an allergic reaction to an insect sting or other immunogenic stimuli, or the use of certain drugs (Table 52.4).383,384,385,386,387,388,398,400,401,402,403,404,405,406,407,409,410,411,415 In both adult and pediatric patients, malignancies, especially Hodgkin disease, have been associated with the development of MCNS (see section on Nephrotic Syndrome and Malignancies).
Clinical Findings
Edema formation may begin within a few days of the inciting event. Facial edema usually is noted first, with few other indications of an ongoing disease process. This can be confused with allergic symptoms, especially if associated with an upper respiratory tract infection. Edema usually increases gradually. It becomes detectable in the adult only when several liters of fluid have accumulated; by the time medical advice is sought, the patient typically has pitting edema involving the sacrum and the lower extremities. When anasarca is present, periorbital edema can be so severe that the eyelids are swollen shut, scrotal or vulval edema may be marked, and there may be significant abdominal distension. Respiratory embarrassment may occur from accumulation of either pleural or ascitic fluid, although the infrequency of dyspnea or orthopnea in the setting of massive fluid retention is striking. This reflects the absence of increased pulmonary capillary wedge pressure needed to generate pulmonary edema. Headaches and irritability are common accompanying complaints of edema. The patient may note vague symptoms such as malaise, easy fatigability, irritability, and depression. Rarely, the development of cellulitis, peritonitis, or pneumonia may be the first indication of an underlying nephrotic syndrome. The pallor resulting from edema can be misinterpreted as indicating anemia.
TABLE 52.4 Factors Reported to Have Precipitated Minimal Change Nephropathy (MCN) | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
On physical examination, dependent edema is the most prominent finding. The retina has a characteristic “wet” appearance. Subungual edema may reverse the usual color pattern on the fingernails—the normally white lunulae may be pink and the rest of the nail bed white. Horizontal white lines that may be seen on both the fingernails and the toenails are referred to as Muehrcke bands. Inguinal and umbilical hernias may be present, especially if the patient has had severe ascites for a prolonged period. The elasticity of the cartilage in the ear appears to be decreased.
Blood pressure in patients with MCN usually is normal, but elevated systolic pressure was recorded in 21% and elevated diastolic pressure in 14% of the children evaluated by the International Study of Kidney Disease in Children (ISKDC).416 Hypertension is seen more commonly in adult patients with MCN.417
Growth failure occasionally may be found in children, most often in those who have had multiple relapses of MCNS.418 Evidence for infection, especially peritonitis, cellulitis, or pneumonia, should be sought as part of the physical examination. These infections may be associated with septicemia and shock.
In MCNS, the chest radiograph usually shows a small or normal-sized heart; pleural effusions may be present, as may pneumonic infiltrates. The presence of an increased heart size and congestive changes in the hilar regions suggests that the nephrotic syndrome is secondary to glomerulonephritis.
Laboratory Findings
Urinalysis
Clinicians who first characterized the nephrotic syndrome noted that the urine often foams excessively when voided and that it coagulates when heated. These findings result from marked proteinuria, now indicated by a dipstick reading of 3+. Other edema-forming, hypoalbuminemic states, such as malnutrition, milk protein sensitivity, and protein-losing
enteropathy, can mimic nephrotic syndrome but do not manifest significant proteinuria. The amount of protein in the urine of nephrotic patients can range from less than 1 g to more than 25 g per day. The value for adult patients usually is between 3.5 and 16 g per day; that for children typically is lower than this amount, even when allowances are made for body size419 and averages about 50 mg per kilogram of body weight per day.420 Because urine protein is a function of plasma, and thus of filtrate, protein concentration,421 children with MCNS, who may have serum albumin levels of 1 g per deciliter or less, may occasionally have amounts of urinary protein as low as 100 to 200 mg per day. This finding in patients with low plasma albumin concentrations also reflects the removal of much of the protein from the glomerular filtrate as it traverses the proximal convoluted tubule422 (see previous section on Tubular Handling of Protein in Mechanisms for Proteinuria). As a consequence of proteinuria, the urine specific gravity in nephrosis usually is high, often exceeding 1.035. Exceptions include patients who are not in an edema-accumulating state (see section on Consequences of Proteinuria, earlier in this chapter) and the patient with nephrotic syndrome and renal failure or tubular dysfunction, in whom lower (but not isosthenuric) values of urine specific gravity are found. Physiologic responses of the kidney to the nephrotic state may cause further urine concentration.
enteropathy, can mimic nephrotic syndrome but do not manifest significant proteinuria. The amount of protein in the urine of nephrotic patients can range from less than 1 g to more than 25 g per day. The value for adult patients usually is between 3.5 and 16 g per day; that for children typically is lower than this amount, even when allowances are made for body size419 and averages about 50 mg per kilogram of body weight per day.420 Because urine protein is a function of plasma, and thus of filtrate, protein concentration,421 children with MCNS, who may have serum albumin levels of 1 g per deciliter or less, may occasionally have amounts of urinary protein as low as 100 to 200 mg per day. This finding in patients with low plasma albumin concentrations also reflects the removal of much of the protein from the glomerular filtrate as it traverses the proximal convoluted tubule422 (see previous section on Tubular Handling of Protein in Mechanisms for Proteinuria). As a consequence of proteinuria, the urine specific gravity in nephrosis usually is high, often exceeding 1.035. Exceptions include patients who are not in an edema-accumulating state (see section on Consequences of Proteinuria, earlier in this chapter) and the patient with nephrotic syndrome and renal failure or tubular dysfunction, in whom lower (but not isosthenuric) values of urine specific gravity are found. Physiologic responses of the kidney to the nephrotic state may cause further urine concentration.
The spectrum of excreted proteins depends on the renal disease responsible for the nephrotic syndrome. In primary nephrotic syndrome, most of the urine protein is albumin; in other diseases, such as glomerulonephritis, both albumin and globulins are lost in increased amounts. This occurrence has led to determination of “protein selectivity” being proposed as a noninvasive method to separate MCNS from other causes of nephrotic syndrome.423 By comparing clearance of albumin to that of larger molecules such as IgG or transferrin, a curve can be generated, indicating whether the protein loss is selective and restricted to small molecules, or nonselective, consisting of both large and small molecules. Patients with MCNS tend to show more selective proteinuria, whereas those with nephrotic syndrome from other causes have nonselective proteinuria. Unfortunately, this generalization is limited by considerable overlap in results from patients in different diagnostic categories, so that the clinical determination of protein selectivity has limited value for individual patients. This limitation may reflect factors other than molecular size that also affect entry of proteins into the glomerular filtrate, and differences in tubular function, which modify the reabsorption of filtered protein.424 A refined electrophoretic technique has been used to indicate protein selectivity and to predict outcomes. Patients with primarily albumin and transferrin in urine as determined by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) proved to be steroid sensitive, whereas patients who were steroid resistant also excreted considerable amounts of IgG, lysozyme, and other larger molecules.425 Some of these larger molecules could be derived from tubular cells, reflecting tubulointerstitial injury rather than glomerular filtration, or from activation of the shunt pathway for filtration of macromolecules discussed earlier in this chapter. In analbuminemic rats with PAN nephrosis, fractional clearances of various macromolecules are similar to those in normoalbuminemic rats. The absence of competition for albumin implied by this finding suggests that the urinary loss of these proteins in nephrotic patients does not result from “overload” of tubular reabsorption by filtered albumin.76
The urine sediment from nephrotic subjects often contains oval fat bodies. Lipiduria is better diagnosed, however, using a microscope equipped with polarized light to demonstrate doubly refractile fat bodies (“Maltese crosses”) in degenerative fatty vacuoles in the cytoplasm of desquamated renal epithelial cells or free in the urine as neutral fat droplets. Frequently, urine with large amounts of protein also contains hyaline casts.
Other urinary findings vary with the cause of the disease. Up to one third of patients with MCN may have microscopic hematuria. Gross hematuria may be seen in patients with uncomplicated MCN but is extremely rare.426 By contrast, it is more common in patients with prominent mesangial proliferation. Hematuria is more likely to be seen with FSGS than MCN.427 In patients with nephrotic syndrome secondary to glomerulonephritis, the urine shows more abnormalities, with cellular elements and granular, cellular, and mixed hyaline casts being present. However, patients with MCN cannot always be differentiated from those with glomerulonephritis on the basis of urine sediment abnormalities alone.
Blood Studies
Hypoproteinemia is common to all nephrotic patients and is caused, primarily, by hypoalbuminemia. Total serum protein is characteristically reduced to between 4.5 and 5.5 g per deciliter; serum albumin concentrations usually fall to below 2 g per deciliter, and, in children, may be less than 1 g per deciliter.379 Although serum albumin concentrations are usually decreased, those of total globulins are remarkably well preserved in MCN despite massive proteinuria. Typically, serum α1-globulin concentrations are normal or slightly decreased, whereas levels of serum α2– and β-globulins are increased. Although the concentration of γ-globulin determined by electrophoresis is normal or reduced, the levels of individual components vary. In MCN,310,403 serum IgG levels average approximately 20% of normal, whereas IgA levels are less severely reduced; IgM and in some cases IgE levels are increased. The changes in serum IgG and IgA concentrations are less pronounced in patients with nephrotic syndrome of other causes; IgM and IgE are typically normal in these subjects.310,403
Hyperlipidemia is one of the findings that define the nephrotic syndrome. Serum total cholesterol level is usually elevated, especially when the serum albumin level has fallen to 2 g per deciliter or less.175 Values average 400 mg per deciliter, but levels in excess of 1,000 mg per deciliter have been recorded. Other changes in plasma lipids are summarized in the section on Consequences of Proteinuria.
Most often, serum electrolyte concentrations are within the normal range even when anasarca is present, indicating a proportionate retention of sodium and water. Factitiously low serum sodium concentrations (˜130 mEq per liter) may be measured with marked hyperlipidemia. This pseudohyponatremia results from the nonaqueous, nonsodiumcontaining component of the serum or plasma (lipid) being increased. It does not require treatment, because the sodium concentration in the aqueous phase of blood is normal, as is plasma osmolality. This artifact is not observed when sodium levels are determined by techniques that measure sodium activity with ion-specific electrodes or after sample ultracentrifugation. Low serum sodium may be an accurate finding in the case of excess free water ingestion relative to dietary sodium intake,428 compounded by potential effects of elevated plasma vasopressin.429 This problem may be exacerbated by diuretic therapy. A decreased anion gap is associated with decreased total serum protein or albumin levels. This finding is common to all hypoalbuminemic states and does not directly reflect either renal dysfunction or altered serum lipid levels.430 Serum calcium may be low, mainly as a result of the hypoalbuminemia. Normally, 40% of total serum calcium is bound to protein. A decrease in serum albumin concentration of 1 g per deciliter results in a decrease in total serum calcium of 0.8 mg per deciliter. In contrast, 1 g of globulin binds only 0.16 mg of calcium. In some cases, the hypocalcemia may be out of proportion to the hypoalbuminemia and is caused by a reduction of ionized calcium levels431 by as much as 5% to 20%, possibly because of urinary loss of 25-OH-D (see section on Consequences of Proteinuria). Acute symptoms of hypocalcemia rarely occur. Total and ionized calcium levels return to normal with remission. Serum phosphorus is normal unless the nephrotic syndrome is associated with renal insufficiency.
Blood urea nitrogen (BUN) and serum creatinine values are usually close to normal in MCN, but may be mildly elevated if decreased intravascular volume from nephrosis causes prerenal azotemia. The BUN may be elevated because of either increased intrarenal urea circulation or increased protein catabolism if the patient has received steroids. The GFR measured by inulin clearance is reduced to an average of 80% of normal157; occasionally, values are reduced to 20% to 30% of normal. This may represent decreased renal perfusion secondary to hypovolemia. Reduced GFR at the onset of MCN is reversible and does not imply an unfavorable outcome.110 The presence or absence of azotemia therefore cannot be used as a reliable indicator in the differential diagnosis of the nephrotic syndrome.
Hemoglobin levels and hematocrit values may be normal or even increased if there is hemoconcentration secondary to a loss of fluid into the peripheral tissues. This factor may help to differentiate azotemic patients with MCN from those who have severe renal insufficiency from parenchymal disease, in whom anemia is more typical. However, as noted previously, iron deficiency may cause nephrotic patients with normal renal function to become anemic.
Measured concentrations of serum complement and its components are generally considered to be normal in MCN. Although urinary losses cause decreases in low-molecular-weight complement components, serum concentrations of the components measured to detect activation of the complement cascade are unchanged.432 Thus, reduced levels of the third component of complement (β-1-C globulin; C3) or C4 indicate that a glomerulonephritis underlies the nephrotic syndrome; conversely, such changes do not occur invariably with glomerulonephritis. Circulating immune complexes may be elevated in MCN or in FSGS.433,434 Plasma renin activity may be increased in some patients who manifest physiologic changes consistent with decreased intravascular volume.
Histopathology
Minimal Change Nephropathy
The morphologic classification of nephrotic syndrome in childhood derives from classic papers by Churg and colleagues435 and White and colleagues.436 The term minimal change nephrotic syndrome was used to describe the pathologic appearance on light microscopy of biopsies from nephrotic patients in which there are no definitive changes from normal in glomeruli (Fig. 52.7). Here, we have chosen to use the term minimal change nephropathy because some adult patients have been reported with proteinuria but no nephrosis9,10 and in order to describe a histologic entity in parallel to FSGS. The degree of change from normal histology that is considered significant remains the subject of some controversy. The spectrum of these changes is classified in Table 52.2 at the beginning of this chapter. Other terms that have been used to describe this general entity include nil disease and steroid-sensitive nephrotic syndrome. Changes in the proximal tubule cells reflect increased reabsorption of protein. Tubular cells may contain apparent vacuoles that are doubly refractile and are similar to the fine lipid droplets seen in oval fat bodies in the urine. This pathologic abnormality generated the term lipoid nephrosis, in which there is no tubular atrophy and the renal interstitium is normal. Older patients may show some globally fibrosed glomeruli with associated nephron loss. This finding is rare in children and should not involve more than 5% to 10% of glomeruli, even in elderly patients.3 Staining for glomerular polyanion with Alcian blue, colloidal iron, or ruthenium red may be reduced in the glomerular tufts. No immunoglobulin or complement deposition is observed by immunofluorescence. Electron microscopy (Fig. 52.7) reveals only glomerular epithelial cell foot-process effacement (see next paragraph).437 In some cases this finding may be visualized by high-resolution light microscopy.438 The diffuse effacement of the podocytes that often contain protein-reabsorptive droplets typically results in the appearance of an almost continuous layer of cytoplasm on the urinary side of the GBM. Epithelial cells may appear detached in segments, producing denuded areas of the GBM. The GBM itself, however, appears normal. There
are no electron-dense deposits adjacent to the GBM.439 Historically, 65% to 85% of children with primary nephrotic syndrome have this lesion,440 compared to a prevalence of about 30% of primary nephrotic syndrome and 15% of all nephrosis in adults.9
are no electron-dense deposits adjacent to the GBM.439 Historically, 65% to 85% of children with primary nephrotic syndrome have this lesion,440 compared to a prevalence of about 30% of primary nephrotic syndrome and 15% of all nephrosis in adults.9
 FIGURE 52.7 Findings on renal biopsies from three children with the clinical features of minimal change nephropathy. A: Light microscopy of a patient with MCN. Portions of two glomeruli are shown. Cellularity is normal and the capillary loops are patent. Tubular and interstitial structures are normal in appearance. (Magnification for all light microscopy ×350.) B: Electron microscopy from the same patient. The endothelial cells (En) lining the capillary loop show a normal fenestrated structure; the glomerular basement membrane (GBM) is uniform in thickness and structure. The epithelial cell (Ep) layer shows characteristic fusion of the epithelial foot processes, with the podocytes being in continuous contact with the GBM. Proteinaceous material and a nucleated cell are present in the capillary lumen (CAP). C: Light microscopy in a patient with mesangial hypercellularity. The tubular and glomerular capillary structures are normal, but an increased number of nuclei are present in the mesangial areas of the glomeruli. Immunofluorescent microscopy was negative for immunoglobulins and C3.The patient behaved clinically as one with MCNS. (Histology courtesy of Dr. John M. Kissane.) |
Podocyte Foot Process Effacement
Podocyte foot process effacement is due to retraction, widening, and shortening of foot processes and is not due to fusion.441 With complete effacement, the GBM is covered by thin, sheetlike processes of podocyte cytoplasm, with gaps between the processes of adjacent cells (where protein filtration presumably occurs). Foot-process effacement is associated with a mild reduction in GFR, which is caused mainly by a reduction in the ultrafiltration coefficient Kf (see section on Physiologic and Anatomic Factors Affecting GFR in Consequences of Proteinuria, earlier in this chapter).
Podocyte effacement is accompanied by striking morphologic changes in the cytoskeleton. The continuous layer of podocyte cytoplasm that overlies the GBM shows an increase in microfilaments and the appearance of a dense cytoskeletal band located within the basal portion of the podocyte, adjacent to the GBM. The cytoskeletal band has regions of high density at regular intervals. En face views demonstrate that the filaments are distributed radially from these central densities, suggesting that they may function to distribute mechanical strain and thereby prevent glomerular capillary expansion.442 Endlich et al.443 exposed cultured podocytes to biaxial cyclic stress in order to model stress that might be experienced by podocytes in glomerulomegaly. Transverse stress fibers disappeared and were replaced by radial stress fibers connected to a single actin-rich center. The stress fibers were composed of myosin II, α-actinin, and synaptopodin. These findings differ in important ways from podocytes in vivo, where the actin-rich center is absent and actin is confined to the foot processes. Nevertheless, the results do suggest that podocyte response to stress consistently involves formation of radial cytoskeletal structures.
It appears likely that podocyte foot-process effacement arises by different mechanisms in different settings. Certain podocyte injury models (protamine infusion, reactive oxygen species infusion, PAN administration) are associated with a redistribution of α-dystroglycan from the basal surface of the podocyte, a process that occurs within as little as 15 minutes.444 MCN but not FSGS is associated with loss of podocyte dystroglycan expression.445 Kojima and Kerjaschki446 have suggested that polycations compete with GBM laminin for binding to its receptor, dystroglycan; free dystroglycan is then internalized into podocyte endosomes. This process is dependent upon cellular ATP and participation of the actin cytoskeleton.
Related but distinct mechanisms may explain foot process effacement in FSGS, in particular, adaptive FSGS (see further, the section on Histopathology of FSGS, later in this chapter). Shirato441 and Kriz et al.447 have proposed that the podocyte supports the essential but contradictory functions of structural stability and leakiness (hydraulic conductivity). Cytoskeletal hypertrophy represents cellular adaptation to increased stress. The source of stress might be glomerular enlargement, increased glomerular PGC (associated with the overload state), or increased GBM distensibility (due to GBM damage). Johnson450 has pointed out that the transmembrane oncotic pressure gradient influences the net filtration pressure (PUF) and thereby might also contribute to net hydraulic stress. PUF is the difference between transmembrane hydraulic pressure and the transmembrane oncotic pressure gradient. In the setting of nephrotic-range
proteinuria, plasma albumin is reduced, which reduces the transmembrane oncotic pressure gradient across the capillary wall and thereby increases PUF In the face of mechanical or hydraulic stress, the capillary wall has three defenses: the mesangial cell, the podocyte, and the GBM. The mesangial cell may undergo proliferation and hypertrophy. Little is known about how the GBM responds to stress. The podocyte cannot proliferate and adapts by elaborating a more complex cytoskeleton to defend the structural integrity of the capillary wall. This occurs, however, at the cost of a reduction in total slit diaphragm width, podocyte effacement, and reduced Kf.
proteinuria, plasma albumin is reduced, which reduces the transmembrane oncotic pressure gradient across the capillary wall and thereby increases PUF In the face of mechanical or hydraulic stress, the capillary wall has three defenses: the mesangial cell, the podocyte, and the GBM. The mesangial cell may undergo proliferation and hypertrophy. Little is known about how the GBM responds to stress. The podocyte cannot proliferate and adapts by elaborating a more complex cytoskeleton to defend the structural integrity of the capillary wall. This occurs, however, at the cost of a reduction in total slit diaphragm width, podocyte effacement, and reduced Kf.
An alternative consideration in FSGS is how a loss of specific podocyte structural proteins impairs cytoskeletal integrity, initiating foot-process effacement. This issue was discussed in the section on Molecular Mechanisms of Podocyte Effacement earlier in this chapter.
Variations in Histopathology
Immunoglobulin Deposition. Some patients with all the clinical features of MCN may have minor morphologic differences from those already described. A common variation is the presence of IgM in the glomerular mesangium. An early report448 suggested that this variant represents a separate entity, which was termed mesangial IgM nephropathy. All of the patients, whose ages ranged from 1.5 to 59 years, showed a slight increase in the mesangial matrix, and in addition to the IgM deposits, some had C3 and rare IgA deposition. Dense mesangial deposits were noted by electron microscopy in 9 of the 12 subjects. Subsequent observations did not support this as being a separate entity. In one study, 40% of 149 consecutive patients with the clinical picture of primary nephrotic syndrome were found on biopsy to have mesangial IgM deposits. Of these, 20 had mostly or entirely IgM without complement. They could not be differentiated clinically from other MCN patients.449 Because the presence of a mesangial IgM deposit in patients with clinical MCN does not appear to affect either the patient’s response to treatment or the disease outcome,450 it is now believed that this finding has little significance.
Mesangial IgM deposits in apparent MCN may, in some cases, be associated with immune complexes.451 Because deposits are often found in patients who undergo biopsy after receiving a trial of corticosteroid therapy, it is of interest that experimental models of immune complex metabolism suggest that steroid administration may prolong the systemic half-life of larger complexes and increase and sustain their appearance in the mesangium.452 The presence in the glomeruli of immunoglobulins in addition to IgM usually indicates a diagnosis of a disease other than MCN.453 One group of patients with a clinical diagnosis of MCN showed some glomerular proliferation associated with immunoglobulin deposits, primarily IgG, in the glomeruli. This lesion was more often observed in African American children,125 whereas a racial predilection may not be present in adults.454 Although the patients described in this report responded to treatment with steroids initially, their subsequent course was one of frequent relapses or the development of resistance to treatment.
Mesangial Proliferation. Some patients with otherwise typical MCN have increased numbers of mesangial cells in the glomeruli. One study that correlated glomerular morphometry with the patient’s clinical course found increased numbers of mesangial nuclei and smaller nuclear sizes in patients who had frequent relapses. The authors proposed that this indicated mesangial cell activation. They cautioned, however, that disease duration could play a role in the development of this finding, because the frequently relapsing patients had a 4-year course compared to 1.4 years in the population with infrequent relapses.455 Mesangial hypercellularity may be associated with a decreased response to steroid therapy,143,456,457 frequent relapse,456 steroid dependency,458 or a poorer prognosis.458,459,460 The ISKDC found that approximately 2.5% of children with nephrotic syndrome had mesangial hypercellularity.461
Mechanisms of mesangial proliferation. Based on the finding of identical immunohistochemistry in patients with or without mesangial proliferation, it has been argued that mesangial IgM deposition does not appear to play a role in the induction of the mesangial cell response.457 Intrinsic kidney cells and cells migrating into the kidney as part of the inflammatory response release factors that regulate mesangial cell proliferation. Mesangial cells produce plateletderived growth factor (PDGF), a stimulant of mesangial and endothelial cell growth and wound healing.462,463,464 Prostacyclin and thromboxane, produced by a variety of cells, are stimulatory cofactors for mesangial cell proliferation.465 In addition, two autocrine growth mechanisms have been defined. The first involves interleukin (IL)-1. Mesangial cells in culture secrete a mesangial cell growth factor with characteristics identical to those of IL-1.466 Indeed, these cells express messenger RNA (mRNA) for IL-1 in vivo.467 The second autocrine system involves IL-6. Mesangial cell-derived IL-6 stimulates mesangial cell growth in vitro.468 Moreover, mice transgenic for the human IL-6 gene show marked mesangial proliferation.469 In human disease, urinary excretion of IL-6 and mesangial staining of biopsy material for IL-6 were associated with mesangial proliferation by some,470 but not all471 authors. Because mesangial proliferative changes are associated with steroid resistance, it is noteworthy that IL-6-induced cell activation is not inhibited by steroids.472 Finally, negative regulation of mesangial cell growth may be provided by the GBM itself, because the proliferation of mesangial cells is decreased by heparin sulfate.473
Tip Lesion. A group of patients has been described as having steroid-responsive nephrotic syndrome with intercapillary foam cells adherent to the Bowman capsule in a tuft near the tubular origin (the glomerular tip lesion). Although this adhesion is irreversible, the patients appear to have a good prognosis closer to that of MCN than that of
steroid-resistant nephrotic syndrome.10,474 Tip lesion is considered more extensively in the section on Selected Clinical Variants of FSGS.
steroid-resistant nephrotic syndrome.10,474 Tip lesion is considered more extensively in the section on Selected Clinical Variants of FSGS.
Disease Processes and Other Findings Associated with MCN
Several clinical findings have been associated with MCN, including specific malignancies, atopy, various human leukocyte antigen (HLA) haplotypes, and abnormalities in immune function. These associations, which could elucidate issues of both causality and treatment, will be considered here.
Malignancy. Several glomerulopathies, notably membranous glomerulopathy, have been associated with neoplasia. A significant number of patients with cancer-related nephrotic syndrome, however, have MCN. The relationship between MCN and lymphomatous disorders, particularly Hodgkin disease, is especially striking.402 In a survey of the literature, 33 of 134 patients with cancer-related nephrotic syndrome had MCN, as determined by biopsy.401 Of the patients with MCN, 26 had Hodgkin disease and an additional two had non-Hodgkin lymphoma. In another review,400 36 of 44 patients who had Hodgkin disease and the nephrotic syndrome had MCN and only 2 had membranous glomerulonephritis. There was a much higher incidence of nephritic diseases in patients with other types of neoplasia. Nonlymphomatous tumors that have been associated rarely with MCN include renal oncocytomas,475 embryonic cell tumors,401 pancreatic carcinoma,476 nephroblastoma,477 Waldenström macroglobulinemia,478 bronchogenic carcinoma,479 and cecal adenocarcinoma.480
Evidence suggests that in these cases the tumor may be directly involved in the pathogenesis of the MCN. MCN can be the initial presenting sign of a lymphomatous disorder481 and may precede clinical evidence of lymphoma by several years.482 With appropriate and successful antineoplastic therapy, either medical or surgical, the proteinuria in tumor-related MCN resolves, renal function remains normal, and the nephrotic syndrome remits.401,475,479,482 The relationship between relapse of the tumor and of the nephrotic syndrome483,484 also strongly suggests an etiologic role for the tumor in the pathogenesis of MCN in these patients. These observations indicate the importance of considering a malignancy as an underlying cause in any adult483,484,485,486 and, rarely, in pediatric487 patients who present with apparent primary nephrotic syndrome. If Hodgkin lymphoma is the cause, it is essential to treat the neoplasm rather than the renal disease.
Atopy and Minimal Change Nephropathy. A relationship between allergy and MCN has long been suspected. Anecdotal reports suggest that exposure to allergens may precipitate the nephrotic syndrome393,394,487,488; rhinorrhea and allergic skin reactivity frequently precede relapses,489 and a high prevalence of allergic symptoms has been observed in nephrotic patients. Highly allergic patients who had a pathologic diagnosis of MCN based on renal biopsy experienced a decrease in urinary protein excretion when placed on an elemental diet and did not require treatment with corticosteroids. Challenge with milk led to a decrease in serum C3 and increased protein excretion, strongly suggesting that hypersensitivity was causally linked to proteinuria.395 A human basophil degranulation test was positive in 16 of 28 adults with MCN and 14 of 18 adults with FSGS; in contrast, only 5 of 29 patients with glomerulonephritis and 1 of 11 healthy donors showed a positive response.490 In addition, atopy and MCN were associated with increased frequency of expression of human leukocyte antigen (HLA)B12 and -DRw7.490,491,492
Although values ranged widely, mean serum IgE levels were significantly elevated in patients with MCN compared to those with other renal problems493; elevated levels also are associated with frequent relapse in children.494 In one study, the majority of adult MCN patients had serum IgE levels more than two standard deviations above the normal mean; of these, more than 70% had associated allergic symptoms.495 Other investigators have made similar observations, but sought to draw a distinction between primary allergic disease and the elevated IgE levels found in nephrotic syndrome. They suggest that because IgE deposition in the glomerulus is rare,496 elevated serum IgE levels could represent not the causal factor in MCN but rather evidence of more generalized derangement in the immune system.497 A finding that increased serum IgE may persist even with remission supports this concept. If this view is correct, it could explain why attempts to treat MCN in atopic patients with inhaled disodium cromoglycate498 or an orally administered analog489 were unsuccessful.
These apparently conflicting observations may be resolved by the study of specific antigens. Therefore, a majority of adult nephrotic patients studied by Meadow et al.,489 Lagrue and Laurent,499 and Laurent et al.500 had detectable elevations of specific IgE titers, with the most common sensitizing agent being house dust or dust mites. After remission, which was induced by the institution of specific desensitization and sodium cromolyn, several of these patients had a relapse on reexposure to the allergen.489,499,500
Immunogenetics. There may be a familial incidence of MCN. A survey from Europe, which excluded patients with the congenital nephrotic syndrome,501 found that 63 of 1,877 nephrotic children had affected family members. This prevalence of 3.35% was higher than that predicted from the frequency with which MCN occurs in the general population. Siblings were most often affected. The similarity of pathologic findings and the clinical course for affected members within a family was striking,502 although in at least one report, siblings showed differences in these features.503 Familial nephrotic syndrome in children has been divided into two broad categories: (1) patients with an infantile onset and a
poor prognosis regardless of renal morphology and (2) patients with a juvenile onset and a generally good response to conventional therapy, provided MCN is found by renal biopsy.
poor prognosis regardless of renal morphology and (2) patients with a juvenile onset and a generally good response to conventional therapy, provided MCN is found by renal biopsy.
An indication of a possible genetic predisposition for the development of MCN is the reported association of MCN, and in some cases atopy,492,504 with certain histocompatibility-complex antigens. The most commonly cited are HLA-B8492,504,505,506 and -DRw7.507,508 Not all studies have confirmed these associations. Indeed, a variety of HLA antigens have been associated with MCN491,492,505,506,507,508,509,510,511,512,513,514,515 and negative associations have been reported, too. HLA-B8 was found frequently in families with more than one member having childhood nephrosis.516 In one study, the combined occurrence of B8 with DR3 and DR7 produced a relative risk of 21.5.514 In another study, DR7 was linked to steroid-sensitive disease and DR3 to steroid-resistant disease.510 In French and German patients, specific DQB1 alleles have been associated with MCNS, most strongly in steroid-dependent or frequently relapsing patients.517 German studies also suggested that patients with HLA-DR7/-DR3 together are less steroid sensitive,518 and those with HLA-DR7 are less likely to respond permanently to alkylating agent treatment of frequent relapse.519
MCN was associated with HLA-DQB1*0601, -DRB1*01, and -DRB1*07011 in Egyptian children509,520 and -DQA*0201 in German children.511 Studies from Japan linked steroid-sensitive MCN in adult patients to HLA-DRw8 and -DQw3512 and to specific DQB1 alleles.521,522 Different results obtained in Singaporean523 or Bengali524 children are likely due in part to racial or geographic differences. The observation that two extended HLA haplotypes (HLA-A1, -B8, -DR3, -DRw52, SCO1; and HLA-B44, -DR7, -DRw53, and FC31) occurred with higher than expected frequency in children with steroid-sensitive, frequently relapsing MCN provided strong evidence for an immunogenetic predisposition to the disease.513
The association between HLA type and MCN has been made primarily in children, with some studies being unable to make similar correlations in adult patients. This finding suggests that MCN in adult and pediatric patients could represent different diseases that share common pathologic and clinical features. For example, HLA-DR7, which has been linked to MCN, was observed in only 18% of adult European MCN patients, a frequency not different from that of control subjects. If the data were analyzed according to age at onset of the nephrotic syndrome, 45% of patients in whom onset was before the age of 15 years were HLA-DR7, whereas the equivalent incidence in adult-onset patients was only 7%.525 It remains unclear whether these associations represent linkage disequilibrium with another gene or reflect potential underlying immune influences on the development and response of MCN.
Disordered Immunity in Minimal Change Nephropathy. It has long been recognized that immunogenic stimuli may precipitate the presentation or relapse of MCN. In addition to the frequent association with atopy (discussed previously), episodes may follow upper respiratory tract infections, bee stings, or diseases linked with abnormal immune responses. The relationship of immunogenic events to the onset of disease and the finding of disordered immune responses in these patients led Shalhoub526 to propose a unifying hypothesis relating MCN to the immune system. He cited four points: (1) evidence for abnormal humoral immune responsiveness; (2) marked sensitivity of the disease process to treatment with corticosteroids or immunosuppressive agents; (3) remission of MCN upon infection with measles, an inhibitor of cell-mediated immunity; and (4) the association of MCN with Hodgkin disease. In view of the lack of significant morphologic evidence of renal damage, Shalhoub suggested that MCN represents the renal manifestation of a systemic immunologic abnormality, perhaps a T-lymphocyte disorder, rather than a primary renal parenchymal disorder. Although subsequent investigation has not yet demonstrated a causal immunologic event, numerous abnormalities in both humoral and cellular immune responses527 have been noted in nephrotic patients (Table 52.5).528,529,549 These may, in time, provide a pathogenic mechanism.
Immunoglobulin synthesis. Clinical and in vitro assays show impaired immunoglobulin synthesis in MCN. Serum IgG levels are decreased significantly in children with MCNS, whereas IgM levels are markedly increased.310,532,533,534 These values return toward normal with remission, although the IgM levels may remain elevated. Not all studies have found an equal tendency toward normalization with remission, nor is this abnormality confined to MCN in all cases.575,576 Although nephrotic proteinuria may be associated with urinary loss of IgG,577 such losses are insufficient to explain the very low serum IgG levels often found in MCN.531 This pattern of increased IgM and decreased IgG levels in the serum also is associated with some other immune-deficient states, most notably X-linked immunodeficiency disease.531
Depression of specific antibody titers, such as those to the common streptococcal antigens endostreptosin, streptolysin O, and streptozyme, was observed in children and adults with idiopathic nephrotic syndrome.535 Levels were low during active disease, remained low for up to 20 years afterward, and were not changed by steroid therapy. Patients who were nephrotic from chronic glomerulonephritis, systemic lupus erythematosus, membranous nephropathy, diabetes mellitus, or amyloidosis did not have depressed titers. These data suggest a chronic, specific impairment of response in patients with MCN. An inability to generate311 or to maintain536 specific titers against pneumococcal polysaccharide has been described in MCN, but not all studies confirmed this observation.578 Thus, depression of specific antibody titers may be restricted to certain patients or certain antigens.
Several groups evaluated the in vitro secretion of immunoglobulins by lymphocytes activated with lectins. Consistent with the decrease in serum IgG levels, pokeweed mitogen-stimulated synthesis of IgG by patient lymphocytes
was decreased in MCN patients during the active stage of disease, returning toward normal with remission.531,533,538 Unstimulated secretion of immunoglobulin may be increased,531,537 suggesting spontaneous activation of lymphocytes in MCN. Studies of IgA and IgM synthesis by lymphocytes obtained from patients with nephrotic syndrome of other causes produced more variable results.579 Decreased immunoglobulin production in vitro or in vivo could result from either abnormalities of lymphocytes or the presence of inhibitory agents systemically or on the cell surface. Evidence suggests that both mechanisms may be present in MCN.
was decreased in MCN patients during the active stage of disease, returning toward normal with remission.531,533,538 Unstimulated secretion of immunoglobulin may be increased,531,537 suggesting spontaneous activation of lymphocytes in MCN. Studies of IgA and IgM synthesis by lymphocytes obtained from patients with nephrotic syndrome of other causes produced more variable results.579 Decreased immunoglobulin production in vitro or in vivo could result from either abnormalities of lymphocytes or the presence of inhibitory agents systemically or on the cell surface. Evidence suggests that both mechanisms may be present in MCN.
TABLE 52.5 Immunologic Abnormalities in Minimal Change Nephropathy | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Studies of lymphocyte surface marker expression. These studies were employed to determine whether the immunoglobulin abnormalities in MCN reflect some form of immune cell dysfunction. Cells infiltrating the renal interstitium are predominantly T lymphocytes.580 The ratio of helper cells to suppressor cells in the glomerulus can be similar to that found in lymphocytes in the peripheral circulation580 but may vary from one patient to another.581 Immunostaining detects higher numbers of CD3-positive glomerular and interstitial T cells, but fewer FoxP3-positive cells, in biopsies from nephrotic patients than from controls. In contrast, there was no difference in CD4-positive or CD8-positive cells. Interstitial, CD68-positive macrophages were also higher in nephrotic patient biopsies.582 Circulating lymphocyte subsets in MCN were initially reported to show no significant alterations in helpersuppressor cell ratios.583 Studies of B-cell and T-cell subpopulations also produced conflicting data, regardless of whether patients with MCN or FSGS were studied.538,540,542,584,585 A potential increase in the number of cells coexpressing B-cell and T-cell surface markers has been reported, comparable to findings in X-linked immunodeficiency.541 However, in most studies of lymphocyte subpopulations in primary nephrotic syndrome, there are few significant changes. The meaning of the differences that were found remains to be determined. In general, studies showing alterations in lymphocyte subpopulations may be useful in suggesting the possible presence of immune derangement, but inferences of a potential role for these changes in disease pathogenesis should be made with caution unless corroborated by accompanying functional analysis. Some progress in this direction is provided by reports of two-color flow cytometry indicating that the counts for circulating activated total T cells and suppressor or suppressor/inducer T cells are increased, whereas those for activated helper T cells are decreased, during relapse.539,586 Another study found that populations of activated suppressor-inducer cells and suppressor-effector cells are increased in patients whose nephrosis was sensitive to steroid treatment, accompanied by decreased memory T cells and decreased lymphocyte proliferation in response to tetanus toxoid. Lymphocytes of steroid-resistant patients also had increased suppressorinducer cells, but decreased suppressor-effector cells and increased memory T cells, with increased responses to tetanus toxoid.587
Cellular immunity. In studies of delayed-type hypersensitivity to common antigens, MCN patients in relapse had decreased skin reactivity to purified protein derivative of tuberculin (PPD), Candida, live varicella vaccine, streptokinase-streptodornase, and topical dinitrochlorobenzene.545,546,547 Reactivity returns when the patient enters remission.545 In addition, the lymphocytes of patients with MCN manifest decreased local graft-versus-host activity when injected into rats, a finding that can be normalized by preincubation of the cells with a thymic factor.588 These observations may not be restricted to MCN.585
In vitro studies also showed abnormal cellular responses in nephrotic patients. Lymphocytes from MCN patients, but not from normal control subjects or patients who were nephrotic secondary to proliferative glomerulonephritis, were toxic to cultured renal tubular epithelial cells.543 Lymphocytes from some patients also proliferated on exposure to GBM.544 It is not clear from these reports whether the findings represent a primary process or the result of immunologic sensitization after renal abnormality. Blast transformation of patient lymphocytes was decreased in the presence of control or nephrotic serum,550 returning to normal after entry of the patient into remission.547 Other results showed that MCN is associated with increased concanavalin A-activated suppressor cell activity. This finding demonstrates at least the potential for exaggerated suppressor lymphocyte responses and is not consistently found in other renal diseases.551,589
Evidence for abnormal lymphokine activity. Serum from adult nephrotic patients inhibits leukocyte migration in the presence of renal antigens562 and serum monocyte migration inhibitory factor activity present during relapse of MCN disappears with remission of the disease.543 Furthermore, serum from most patients with MCN as well as some with diffuse proliferative glomerulonephritis is lymphocytotoxic, whereas serum from patients with acute tubular necrosis or urologic disease is not.552 These findings suggest that an immune inhibitory agent, or a series of such agents, is present in the circulation of patients with primary nephrotic syndrome. Sera from nephrotic patients may inhibit the ability of cells to form rosettes,553,554 although this is not specific for MCN, and patient sera do not support in vitro antibodydependent cell-mediated cytotoxicity (ADCC) assays.555 Decreased splenic uptake of radiolabeled complexes was correlated with deficient Fc receptor function in nephrotic patients and could be due to the presence of an inhibitory protein that attaches to cell surfaces.528 Finally, multiple studies demonstrate a suppressive effect of patient sera on blastogenesis by normal lymphocytes.561,590,591,592 The specificity of this phenomenon for MCN varies from one study to another. Efforts to attribute suppressive activity in nephrotic sera to a lymphokine should attempt to exclude the possibility that the observed effects are caused by nonspecific toxicity to immune responses, resulting from the biochemical abnormalities that occur in nephrotic syndrome. For example, the suppressive activity in plasma from nephrotic children segregates in the lipid-rich fraction.593 It could thus be derived from constituents of LDLs and VLDLs that are present in increased concentrations in nephrotic plasma and suppress in vitro cellular immune responses.594,595 However, this finding also is consistent with the migration of an immunosuppressive agent with the lipid-rich fraction. Alternatively, oxidized LDL could itself affect lymphocyte function. In either case, it is noteworthy that a patient with refractory nephrotic syndrome entered remission after LDL apheresis.596
Several studies that partially characterized the suppressive activity suggested the production of a suppressor lymphokine. A heat-stable substance, present in the serum of nephrotic patients, inhibits lymphocyte proliferation. It binds to lymphocytes in the assay system and is not removed by washing.560 One study found a heat-stable inhibitory substance in the plasma of 51 (76%) of 67 children with MCN and 6 of 9 children with FSGS.559 Only 1 sample from 7 patients with membranous glomerulonephritis or 31 healthy adults and children showed similar activity. The factor was toxic to normal lymphocytes and was between 100 and 300 kDa. The presence of tumor necrosis factor,567 IL-4,574,597 and IL-13576 in the serum of some patients may be related to suppressive activity.
Urine and serum samples from children and adults with steroid-sensitive nephrotic syndrome, but not other causes for proteinuria, contain the lymphokine soluble immune response suppressor (SIRS).565 This factor, which inhibits antibody production598 and delayed-type hypersensitivity responses,599 is secreted by patient lymphocytes without a requirement for exogenous stimulating agents. SIRS production thus could account for the suppression of immune responses seen in nephrotic patients. Suppressive activity disappears from the urine after the initiation of corticosteroid therapy but before urinary protein loss decreases significantly. Patient serum activates normal lymphocytes to produce SIRS by a steroid-sensitive process,566 and a regulatory mechanism has been proposed by which CD4+ T lymphocytes from patients in relapse secrete a protein that activates CD8+ T cells to produce SIRS.600 Although the parallel between the sensitivity of SIRS production to steroids and steroid responsiveness of nephrotic proteinuria in patients who produce SIRS is striking, there is no evidence to indicate that SIRS itself causes nephrotic proteinuria. It is, however, a clear marker for steroid-sensitive mechanisms of proteinuria. The means by which SIRS acts on immune responses are not known. Although this issue remains intriguing, little recent progress has been made in this area.
Circulating immune complexes. A variety of glomerular diseases have been associated with soluble circulating immune complexes. The circulating complexes reflect immunoglobulins found in the kidney; patients with MCN had little or no IgG or IgA complexes but did have marked variation with regard to circulating IgM complexes.601 Although circulating
immune complexes have been documented in some patients with MCN or FSGS,530 not all studies have confirmed this observation529 and the relationship of this finding to mesangial immune complexes remains uncertain. A possible reason for varied results is the use of different assay systems.433 In screening studies that employed liquid-phase and solid-phase C1q binding and Raji cell assays, at least one assay was positive in serum from 11 of 14 adults with MCN, 13 of 27 patients with FSGS, and 26 of 55 patients with membranous nephropathy. Prednisone treatment did not affect the prevalence of circulating immune complexes in this study.602 In another report, 17 of 18 MCN patients had IgG immune complexes that did not bind to Clq; in 7 of 9 patients assessed longitudinally, immune complexes disappeared within 6 weeks of entry into remission.434 This temporal relationship and the absence of glomerular IgG in patients with MCN suggest that circulating immune complexes could be a result rather than a cause of the disease. Although they may not cause the disease, some circulating immune complexes could account for the apparently nonspecific presence of IgM in the mesangium of some patients. In support of this is the finding that neutral, or anionic, large complexes show focal to diffuse mesangial localization.603,604 Complexes containing IgM tend to be large. In contrast, lowavidity, polycationic, and small immune complexes tend to deposit in capillary or mesangiocapillary distribution.
immune complexes have been documented in some patients with MCN or FSGS,530 not all studies have confirmed this observation529 and the relationship of this finding to mesangial immune complexes remains uncertain. A possible reason for varied results is the use of different assay systems.433 In screening studies that employed liquid-phase and solid-phase C1q binding and Raji cell assays, at least one assay was positive in serum from 11 of 14 adults with MCN, 13 of 27 patients with FSGS, and 26 of 55 patients with membranous nephropathy. Prednisone treatment did not affect the prevalence of circulating immune complexes in this study.602 In another report, 17 of 18 MCN patients had IgG immune complexes that did not bind to Clq; in 7 of 9 patients assessed longitudinally, immune complexes disappeared within 6 weeks of entry into remission.434 This temporal relationship and the absence of glomerular IgG in patients with MCN suggest that circulating immune complexes could be a result rather than a cause of the disease. Although they may not cause the disease, some circulating immune complexes could account for the apparently nonspecific presence of IgM in the mesangium of some patients. In support of this is the finding that neutral, or anionic, large complexes show focal to diffuse mesangial localization.603,604 Complexes containing IgM tend to be large. In contrast, lowavidity, polycationic, and small immune complexes tend to deposit in capillary or mesangiocapillary distribution.
Other findings related to immunity. Further evidence of a potential role for the immune system in MCN is the possible relationship between this disease and allergic phenomena, already discussed, as well as the unique association between MCN and tumors of immune cell origin. Impaired lymphocyte blast transformation was found in the presence of plasma from patients with MCN and Hodgkin disease486,605; in vitro responses improved significantly after antitumor therapy.606 The strong association of lymphoid tumor, abnormal cellular immune responses, and MCN supports a role for deranged immunity. The nature of this derangement is unclear. Despite the clinical evidence of suppressed immune responses, the underlying abnormality paradoxically may be general immune system activation. In support of such an event, production of a number of specific cytokines and their regulators may be elevated. Increased IL-2 mRNA expression in patient lymphocytes,607 increased IL-8 production in steroid-resistant patients,608 and increased incidence of one polymorphism of IL-1 receptor antagonist609 have been reported. Circulating soluble IL-2 receptor concentrations are increased in MCN.569,570,571,572 Production of gamma interferon (IFN-γ) may also be increased in nephrotic children. Further, although various responses of stimulated patient lymphocytes are decreased, unstimulated cells from patients show, for example, increased immunoglobulin production relative to that of unstimulated control cells. Taken together, these findings suggest that in MCN, the immune system is generally activated, whereas the induction of responses to specific stimuli is impaired.527 Consistent with this hypothesis are the recent data indicating that rituximab, which targets B lymphocytes, rather than the T cells implicated by Shalhoub, is effective in MCN (see section on Treatment of MCN).
Relationship of the immunologic abnormalities to disease pathogenesis. Despite all of these studies, Shalhoub’s hypothesis has not yet been proved. There is strong support, however, for the concept that cellular immunity may be a mediator of proteinuria. Monocytes or macrophages are important in the pathogenesis of some forms of glomerulonephritis610 and in the genesis of proteinuria.611,612 CD34-positive immature lymphocytes from patients with MCN induce proteinuria upon their injection into non-obese diadbetic/severe combined immunodeficiency (NOD/SCID)-immunodeficient mice.613 These studies imply a role for mononuclear cells but do not explain how they may act. One possibility is through release of the lymphokine, vascular permeability factor (VPF), which is produced by activated lymphocytes from some nephrotic patients and which, when injected intradermally, causes increased permeability of vessels to macromolecules.563 This protein, usually referred to by its function as an endothelial cell stimulant called vascular endothelial cell growth factor (VEGF),564 was detected in supernatants of unstimulated cultures of patients’ cells but not those of normal controls.614 A VEGF-like serine protease in patient serum decreased staining for polyanion when used to treat histologic sections of normal glomerular tissue.615 However, the effect of VEGF does not appear to be specific for permeability of albumin, making it an unlikely cause of selective proteinuria. VEGF activates endothelial cells in numerous ways; by causing the cells to “round up,” it disrupts tight cell-cell adhesions, promoting permeability of an endothelial monolayer.616 It also promotes the formation of endothelial fenestrae.617 However, because the glomerular capillary endothelium is not the final barrier to glomerular permeability, it is not clear how increased fenestration would cause albuminuria in primary nephrotic syndrome. Furthermore, a similar substance was described in IgA nephropathy, even in the absence of nephrotic syndrome,618 indicating that VEGF activity could be secondary to renal disease rather than a cause of proteinuria. VEGF production could be enhanced by IL-12 or IL-15.619 Substances produced by T-cell hybridomas derived from the lymphocytes of nephrotic patients,620 or found in culture media conditioned by mononuclear cells from nephrotic children,621 may prove to represent one or more selective permeability factors. One such factor could be IL-8, which is produced by lymphocytes from patients with steroid-responsive nephrosis and increases renal clearance of protein in an ex vivo model.622 However, varied circulating levels have been reported in disease.608,623 IL-2573 and tumor necrosis factor alpha [TNFα]624 production by patient lymphocytes normalized after the patients entered remission. A single case report describes remission of steroid-resistant disease after infliximab anti-TNFα therapy.625 Another potentially significant mediator is IL-13. This Th2 cytokine, which has been associated with allergy,
is found in increased concentrations in serum from children with MCN.626 In an animal model, overexpression of IL-13 caused nephrotic-range proteinuria in rats.627
is found in increased concentrations in serum from children with MCN.626 In an animal model, overexpression of IL-13 caused nephrotic-range proteinuria in rats.627
Despite the absence of proof for Shalhoub’s hypothesis, the indirect evidence of an immunogenic basis for many cases of MCN remains compelling. Onset is often preceded by an immunogenic stimulus. Measles, which induces remission, inhibits lymphokine production but not proliferation by lymphocytes.628 Studies of MCN induced by NSAIDs629 or cimetidine630 indicated that these cases of disease may be associated with abnormal T-cell function. The relationship of disease to altered immunity is particularly striking with regard to suppressor cell activity. A good therapeutic response to cyclophosphamide was associated with decreased suppressor cell activity after treatment,631 although others were unable to confirm this finding.632 Furthermore, as described previously, cellular and humoral immune responses are suppressed in MCN. Thus, it is intriguing that recombinant leukocyte interferon A, an agent that induces production of SIRS, causes nephrotic syndrome with minimal glomerular changes in some patients with T-lymphocyte malignancy633 and that the anthelminthic agent levamisole, which inhibits SIRS activity,598 has been used successfully to treat MCN.634,635 Despite evidence suggesting that cytokines may be involved in promoting vascular permeability,636 the data regarding lymphokines do not address their role as a pathogenic agent in nephrosis and are equally consistent with the interpretation that production of these substances is an epiphenomenon of the derangement that causes albuminuria. In addition, the existence of differences between studies or even within a patient group in a given study suggest that multiple etiologies may exist for MCN, only one (or several) of which may be immunologic.
One explanation for the apparent role of the immune system and the sensitivity of disease to treatments that affect immune responses is that podocytes may have molecular signaling pathways that are analogous to those observed in lymphocytes. The molecule, B7-1, termed CD80 in humans, is found on activated B lymphocytes and monocytes and provides a costimulatory signal necessary for T-cell activation and survival. It also is expressed in the podocytes of patients with MCN in relapse.91 When overexpressed in cultured podocytes, it disrupts the cytoskeleton, interfering with assembly of the slit-diaphragm complex.90 In vivo, its expression in podocytes induces proteinuria in mice. Another immunologic signaling molecule, c-maf-interfering protein (c-mip), was previously found in lymphocytes of MCN patients in relapse. It is expressed in the podocytes of mice treated with lipopolysaccharide, an experimental stimulus for proteinuria, and disrupts podocyte architecture when it is expressed in those cells.637 It also has been detected in Reed-Sternberg cells as well as in the podocytes of patients who have MCN related to Hodgkin disease.638 Together, these results suggest a paradigm in which podocytes express signaling pathways analogous to those previously characterized in immunocytes. Rather than mediating immune responses, their activation is associated with proteinuria, perhaps with a teleogic purpose to clear circulating antigens. Because the podocyte and lymphocyte pathways are similar, the podocytes might respond in parallel to the immune system in the presence of systemic immunogenic stimuli that result from atopy or viral infection. These considerations provide a potential relationship between the immune abnormalities that have been associated with MCN and the pathogenesis of proteinuria in the disease.
Treatment of Minimal Change Nephropathy
Indications for Renal Biopsy
Although most older children and adults will undergo renal biopsy prior to starting treatment, the indications for and benefits of this procedure in patients with primary nephrotic syndrome remain somewhat controversial.639 Children 1 to 9 years old with all the features of MCNS and no atypical features (Table 52.6) may be given a therapeutic
trial of steroids without prior histologic confirmation of the diagnosis. Induction of a complete remission by steroids in such patients is considered to be adequate confirmation of the diagnosis of MCN,440 although other benign conditions may occasionally appear to respond to steroid therapy.640 A histologic diagnosis by renal biopsy is recommended in all patients who present with nephrotic syndrome in the first year of life or after the age of 9 years. Risk-benefit analysis at one point suggested that a biopsy may not be diagnostically superior to a therapeutic trial of steroids, even in adults,641 but the increasing incidence of FSGS in nephrotic adults has led most clinicians to perform a biopsy at the time of presentation. All patients who are steroid-resistant (see Table 52.7 for some commonly used descriptors of nephrotic patients and their responses) should undergo biopsy, although the timing of this decision remains uncertain. Although steroid resistance in children usually is defined by 8 weeks of nonresponse, many clinicians recognize that most children who respond do so by 4 weeks. Therefore, it may be appropriate to perform a biopsy after 4 weeks or less, because those who have not responded by then are more likely to have a different disease that might require an alternative treatment. Other pediatric patients who may require a biopsy include those who have frequent relapses, or who are candidates for therapy with immunosuppressive drugs. Some experts have suggested that any steroid-dependent pediatric patients who continue to be steroid sensitive are, statistically, so likely to have MCN416,461 that biopsy is not indicated. Alternatively, response to steroids may itself be a more important determinant of disease process and outcome than is histopathology.642
trial of steroids without prior histologic confirmation of the diagnosis. Induction of a complete remission by steroids in such patients is considered to be adequate confirmation of the diagnosis of MCN,440 although other benign conditions may occasionally appear to respond to steroid therapy.640 A histologic diagnosis by renal biopsy is recommended in all patients who present with nephrotic syndrome in the first year of life or after the age of 9 years. Risk-benefit analysis at one point suggested that a biopsy may not be diagnostically superior to a therapeutic trial of steroids, even in adults,641 but the increasing incidence of FSGS in nephrotic adults has led most clinicians to perform a biopsy at the time of presentation. All patients who are steroid-resistant (see Table 52.7 for some commonly used descriptors of nephrotic patients and their responses) should undergo biopsy, although the timing of this decision remains uncertain. Although steroid resistance in children usually is defined by 8 weeks of nonresponse, many clinicians recognize that most children who respond do so by 4 weeks. Therefore, it may be appropriate to perform a biopsy after 4 weeks or less, because those who have not responded by then are more likely to have a different disease that might require an alternative treatment. Other pediatric patients who may require a biopsy include those who have frequent relapses, or who are candidates for therapy with immunosuppressive drugs. Some experts have suggested that any steroid-dependent pediatric patients who continue to be steroid sensitive are, statistically, so likely to have MCN416,461 that biopsy is not indicated. Alternatively, response to steroids may itself be a more important determinant of disease process and outcome than is histopathology.642
TABLE 52.6 Features Suggesting that Nephrotic Syndrome Is Caused by a Disease Other than Minimal Change Nephropathy (MCN) | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |||||||||||||||
