Monoclonal Gammopathies: Multiple Myeloma, Amyloidosis, and Related Disorders



Monoclonal Gammopathies: Multiple Myeloma, Amyloidosis, and Related Disorders


Pierre Ronco

Frank Bridoux

Pierre Aucouturier



Monoclonal proliferations of the B cell lineage, often referred to as plasma cell dyscrasias, are characterized by abnormal and uncontrolled expansion of a single clone of B cells at different maturation stages, with a variable degree of differentiation to immunoglobulin (Ig)-secreting plasma cells. Therefore, they are usually associated with the production and secretion in blood of a monoclonal Ig and/or a fragment thereof. An ominous consequence of secretion of monoclonal Ig products is their deposition in tissues. These proteinaceous deposits can take the form of casts (in myeloma cast nephropathy [CN]), crystals (in myeloma-associated Fanconi syndrome [FS]), fibrils (in light-chain [LC] and exceptional heavy-chain [HC] amyloidosis), or granular precipitates (in monoclonal Ig deposition disease [MIDD]) (Table 60.1). They may disrupt organ structure and function, inducing life-threatening complications. In a large proportion of patients with crystals, fibrils, or granular deposits of Ig products, major clinical manifestations and mortality are related to visceral Ig deposition rather than to expansion of the B cell clone. Indeed, except for myeloma CN, which is generally associated with a large tumor mass malignancy, Ig precipitation or deposition diseases often occur in the course of a benign B cell proliferation or of a smoldering or low-mass myeloma. 1

The presence of abnormal urine components in a patient with severe bone pain and edema was first recognized in the 1840s by Henry Bence Jones and William MacIntyre, who described unusual thermal solubility properties of urinary proteins, far later attributed to Ig LCs. To perpetuate this discovery, monoclonal LC proteinuria is often referred to as Bence Jones proteinuria . This term is not appropriate because less than 50% of LCs do show thermal solubility. Renal damage characterized by large protein casts surrounded by multinucleated giant cells within distal tubules was identified in the early 1900s and termed myeloma kidney . This term must, however, be abandoned because CN with acute renal failure may occasionally occur in conditions other than myeloma and because other patterns of renal injury were subsequently found in patients with myeloma. The first of these was amyloidosis, wherein tissue deposits are characterized by Congo red binding and fibrillar ultrastructure. In 1971, Glenner et al.2 showed that the amino acid sequence of amyloid fibrils extracted from tissue was identical to the variable region of a circulating Ig LC, thereby providing the first demonstration that an Ig component could be responsible for tissue deposition. The spectrum of renal diseases due to monoclonal Ig deposition has expanded dramatically with the advent of routine staining of renal biopsy specimens with specific anti-κ and anti-κ LC antibodies, and of electron and immunoelectron microscopy (Table 60.1). These morphologic techniques associated with more sensitive and sophisticated analyses of blood and urine monoclonal components have led to the description of new entities, including nonamyloid monoclonal LC deposition disease (LCDD),3 HC (or AH) amyloidosis,4 nonamyloid HC deposition disease (HCDD),5,6 glomerulopathies with organized microtubular monoclonal Ig deposits,7,8 and proliferative glomerulonephritis with non-organized monoclonal IgG deposits.9 All of these pathologic entities principally involve the kidney, which appears as the main target for deposition of mono clonal Ig components. This is not only explained by the high levels of renal plasma flow and glomerular filtration rate (GFR), but also by the sieving properties of the glomerular capillary wall and by the prominent role of the renal tubule in LC handling and catabolism.10,11

Polymorphism of renal lesions may be due to specific properties of Ig components influencing their precipitation, their interaction with renal tissue, or their processing after deposition. Alternatively, the type of renal lesions may be driven by the local response to Ig deposits, which may vary from one patient to another. That intrinsic properties of Ig components are responsible for the observed renal alterations was first suggested by in vitro biosynthesis of abnormal Ig by bone marrow cells from patients with lymphoplasmacytic disorders and visceral LC deposition12 and by recurrence of nephropathy in renal grafts.13 A further demonstration of the specificity of Ig component pathogenicity was provided by Solomon et al.14 They showed that the pattern of human renal lesions associated with the production of monoclonal LC, that is, myeloma CN, LCDD, and LC (or AL) amyloidosis,
could be reproduced in mice injected intraperitoneally with large amounts of LCs from patients. The good correlation between experimental findings and human lesions led to the conclusion that physicochemical or structural properties of LCs might be responsible for the specificity of renal lesions.








TABLE 60.1 Pathologic Classification of Diseases Featuring Tissue Deposition or Precipitation of Monoclonal Immunoglobulin-Related Material






































Organized


Nonorganized


Crystals


Fibrillar


Microtubular


MIDD (“Randall type”)


Other


Myeloma cast nephropathya


Amyloidosis (AL, AH)


Cryoglobulinemia kidney


LCDD


GN with monoclonal IgG


Fanconi’s syndrome


Nonamyloid


Immunotactoid


LHCDD


Crescentic GN (IgA or IgM)


Other (extrarenal)




HCDD



a Crystals are predominantly localized within casts in the lumen of distal tubules and collecting ducts, but may also occasionally be found in the cytoplasm of proximal tubule epithelial cells.


AH, heavy-chain amyloidosis; AL, light-chain amyloidosis; GN, glomerulonephritis; HCDD, LCDD, LHCDD, MIDD, heavy-chain, light-chain, light- and heavy-chain, monoclonal immunoglobulin deposition disease.


Adapted from Preud’homme JL, Aucouturier P, Touchard G, et al. Monoclonal immunoglobulin deposition disease (Randall type): relationship with structural abnormalities of immunoglobulin chains. Kidney Int. 1994;46:965, with permission.


A normal Ig is composed of two LCs and two HCs, which are themselves made up of so-called constant (C) and variable (V) globular domains. Whereas a limited number of genes encode the constant region, multiple gene segments are rearranged to produce a variable domain unique to each chain. Diversity is further amplified by junctional molecular events that affect the third hypervariable zone (CDR3), and then by the hypermutation process that occurs in the germinal centers of lymphoid follicles. Consequently, although LCs (and HCs) have many structural similarities, they also possess a unique sequence that may be responsible for physicochemical peculiarities, hence their deposition in tissue or interaction with tissue constituents. A number of structural and physicochemical abnormalities of Ig have already been described. They include deletions of CH domains in HCDD5,6 and HC amyloidosis,4 shortened or lengthened LCs and abnormal LC glycosylation in LCDD,12,15 and resistance to proteolysis of the VL fragment in FS.16 Moreover, overrepresentation of certain VL gene subgroups was also reported in amyloidosis17,18 and LCDD.19 The mechanisms generating Ig diversity may randomly create HCs or LCs with peculiar properties such as proneness to deposition, whereas mistakes in the rearrangement or hypermutation processes may result in altered genes encoding truncated Ig. It must be stressed, however, that some abnormal Ig chains produced in immunoproliferative disorders are not associated with any special clinical features. Conversely, structural abnormalities of LCs are not a constant feature of diseases associated with LC deposition. These observations suggest the need to increase the number of nephritogenic Ig components to be analyzed at the complementary DNA (cDNA) and protein levels.

Myeloma- and AL amyloidosis-induced renal failure accounts for less than 2% of the patients admitted to a chronic dialysis program each year.20 This is due in part to the relative rarity of these immunoproliferative diseases, but also to a deteriorated clinical condition of patients at the time of end-stage renal disease. A substantial effort of prevention must therefore be carried out, relying in part on a better understanding of the structural and physicochemical properties of Ig components leading to deposition or precipitation in tissues. Any progress in this field may also enlighten the pathogenesis of immunologically mediated renal diseases, especially glomerulonephritides, because properties of monoclonal Ig components favoring their deposition may apply as well to polyclonal Ig involved in the formation of immune complexes.

We have classified the various forms of renal involvement in monoclonal gammopathies according to the lesions observed in renal biopsy specimens. The majority of patients (63% in a series of 87) with serum and/or urine monoclonal gammopathy who undergo renal biopsy have disease unrelated to monoclonal gammopathy deposition.21 Therefore, the diagnosis of virtually all of the entities to be discussed is critically dependent on the inclusion of κ and λ in the standard of immunofluorescence stains. In some of the rarer entities, a more refined and precise diagnosis can be made with immunofluorescence staining for the subclasses of IgG. Collectively these stains may demonstrate light chain isotype restriction and κ-heavy chain subclass restriction, which strongly favors, but does not definitely prove, the presence of a monoclonal Ig. Demonstration of monoclonality requires serum and urine studies by immunoelectrophoresis or immunofixation.



MYELOMA-ASSOCIATED TUBULOPATHIES

The prevalence of tubular lesions in patients with myeloma is difficult to assess because most patients do not undergo a renal biopsy, but it is most likely high. In Ivanyi’s necropsy study including immunofluorescence, 18 of 57 patients (32%) had CN, whereas 6 (11%) had renal amyloidosis and 3 (5%) had κ-LCDD.22 The higher prevalence of CN (30%) was confirmed in the more recent autopsy series of Herrera.23 Tubular alterations are also demonstrated by increased urinary concentrations of the low molecular weight proteins normally reabsorbed by the proximal tubule, increased urinary elimination of the tubular lysosomal enzyme β-acetyl- D – glucosaminidase, and frequent abnormalities in renal tubular acidifying and concentrating ability24 in patients with LC proteinuria. However, myeloma-associated FS remains an exception.

CN is not only the most frequent lesion in myeloma patients, it is also the major cause of renal failure, which is observed in about 25% of patients with multiple myeloma. In nephrology departments that usually receive only myeloma patients with severe renal abnormalities, the prevalence of CN assessed histologically varies from 63% to 87%25,26,27,28 among the myeloma patients with renal failure. This prevalence is most likely underestimated because patients with presumed CN do not systematically undergo a renal biopsy, whereas those exhibiting significant albuminuria or a fortiori the nephrotic syndrome do. In myeloma patients with an albumin urinary output of less than 1 g per day, there is a good correlation between the diagnosis of CN and renal failure. Of note, CN may occur in other immunoproliferative disorders featuring urinary LC excretion including Waldenström macroglobulinemia29 and µ-HC disease.30 In a case of µ-HC disease, the urinary secretion of large amounts of free κ-chain was responsible for acute renal failure with a typical histologic presentation of “myeloma kidney”.30






FIGURE 60.1 Schematic representation of the pathogenesis of myeloma cast nephropathy. GFR , glomerular filtration rate; THP , Tamm-Horsfall protein. (Adapted from Winearls CG. Nephrology forum: acute myeloma kidney. Kidney Int. 1995;48:1347.)


Myeloma Cast Nephropathy


Pathophysiology of Myeloma Cast Nephropathy

CN occurs mainly in patients with myeloma with a high LC secretion rate. That LCs are the main culprits is supported also by the following clinical, pathologic, and experimental data:

1. Renal lesions may recur on grafted kidneys.

2. Similar crystals may occasionally be seen within casts, proximal tubule cells, and plasma cells. Their usual lack of staining with anti-LC antibody is most likely due to degradation or masking of the relevant epitopes.

3. Mice injected with LC purified from patients with CN developed extensive cast formation in the distal renal tubules.14

However, a number of patients produce large amounts of LCs and yet fail to present significant signs of renal involvement throughout the course of the disease. This may be related to the absence of enhancing factors (see later text), but this also suggests that some LCs may be particularly prone to induce renal lesions, especially cast formation.

LCs are directly toxic to epithelial cells, resulting in decreased proximal reabsorption of the LCs and increased delivery to the distal tubule in which they coprecipitate with Tamm-Horsfall protein (THP). Tubular obstruction by large and numerous casts may also contribute to the development of tubular lesions. For clarity, we will analyze separately the pathogenesis of proximal tubule lesions that result from renal metabolism of LCs, the mechanisms of cast formation, and the respective role of tubular obstruction and tubular lesions in the genesis of renal failure (Fig. 60.1).


Renal Metabolism of Light Chains and Pathogenesis of Proximal Tubule Lesions. Normal as well as malignant plasma cells can secrete free LC, in addition to complete Ig molecules; the amount of free LC secretion is highly variable, depending on the variable ( VL ) domain structure. The LCs are normally filtered by the glomerulus and then reabsorbed by the proximal tubule. Lambda and, to a lesser extent, κ-LCs circulate mainly as covalently linked dimers that have a mass-restricted glomerular filtration. In normal individuals, several hundred milligrams per day of circulating free polyclonal LCs are filtered by glomeruli and more than 90% of these are reabsorbed and catabolized by proximal tubular cells. LCs bind to a single class of low affinity, high capacity noncooperative binding sites on both rat and human kidney brush-border membranes. These sites exhibit relative selectivity for LCs compared with albumin and β-lactoglobulin. It has been shown that LCs could bind to the tandem receptor cubilin10 and megalin,11 a multiligand receptor belonging to the large family of low density lipoprotein receptors, located in the intermicrovillar areas of the brush border. After binding to the luminal domain of proximal tubular epithelial cells, LCs are incorporated in endosomes that fuse with primary lysosomes where proteases, mainly cathepsin B, degrade the proteins into amino acids, which are returned to the circulation by the basolateral route.

When the concentration of filtered LCs is increased as in myeloma patients, profound functional and morphologic alterations of proximal tubule epithelial cells may occur. The functional disturbances include low molecular weight proteinuria and inhibition of sodium-dependent uptake of amino acids and glucose by brush-border preparations. Furthermore, in human proximal tubule cells, endocytosis of LCs was shown to induce activation of redox pathways,31,32 NF-κB,33 and mitogen activated protein kinases (MAPK),34 resulting in cytokine production including interleukin (IL)-6, IL-8, transforming growth factor (TGF)-β, and monocyte chemoattractant protein-1 (MCP-1). Part of these effects may be mediated by megalin itself because megalin possesses intrinsic signaling properties. Moreover, excessive LC endocytosis may promote apoptosis35 and induce epithelial-mesenchymal transition of tubular cells.36,37 Increased cytokine production may be a major mechanism mediating tubulointerstitial injury and progressive kidney disease in some patients with myeloma. Suppression of proinflammatory cytokine production using inhibition of p38 MAPK and translocation of NF-κB by pituitary adenylate cyclase- activating polypeptide with 38 residues (PACAP38) was shown to dramatically prevent injury of cultured renal proximal tubule cells caused by myeloma LCs.38

Morphologically, some of the LCs infused in mice or rats or perfused in rat nephrons in vivo accumulated in enlarged, distorded endosomes and lysosomes of the proximal convoluted tubule with frequent crystalloid formations. This was associated with mitochondrial alterations, focal loss of the microvillus border, and epithelial cell exfoliation.

Pathogenesis of Cast Formation. Because myeloma casts are composed principally of the monoclonal LC and THP/uromodulin, it has long been hypothesized that interaction of these two proteins was a key event in cast formation. THP is a highly glycosylated and acidic protein (isoelectric point [pI] = 3.2) synthesized exclusively by the cells of the ascending limb of the loop of Henle. It is the major protein constituent of normal urine, and an almost universal component of casts. This 80-kDa protein is also remarkable for its ability to form reversibly high-molecular-size aggregates of about 7 × 106 daltons at high but physiologic concentrations of sodium and calcium, and at low urinary pH. The role of THP in cast formation has prompted a wealth of studies on its interactions with LCs. These studies were performed with the aim of defining a population of myeloma patients at risk of developing renal damage. The role of LC pI has long been suggested. It was proposed that LC with a high pI (greater than 5.6) and THP could bear opposite charges in the normal urine pH range, and undergo polar interaction and precipitation. However, this hypothesis was not confirmed in further experimental and clinical studies.25,39

In a rat model, development of casts and injury to proximal tubule cells in renal tubules microperfused with human nephritogenic LC were not correlated with LC pI, molecular form, or isotype.40 Intranephronal obstruction was aggravated by decreasing extracellular fluid volume or adding furosemide. In perfused loop segments, cast-forming LCs reduced chloride absorption directly, thereby increasing tubule fluid [Cl˜] and promoting their own aggregation with THP.41 Pretreatment of rats with colchicine, which prevents addition of sialic acid to the protein, completely prevented obstruction and cast formation in perfused nephrons, and THP from those rats did not aggregate with LCs in vitro, contrary to THP purified from control rats. In vitro studies suggest that THP can undergo both self (homotypic) aggregation and heterotypic aggregation with LCs. Homotypic aggregation is enhanced by calcium, furosemide, and low pH, and is dependent on THP sialic-acid content. Heterotypic aggregation requires previous binding of LC to the THP protein backbone. A 9-residue sequence of the THP was identified as a binding site of LCs, including a histidine at position 226, which explains, at least partially, the pH dependence of molecular interactions.42 LCs bind to THP through their third complementary determining region (CDR3).43 The sugar moiety is also essential for coaggregation of LC and THP. THP from normal volunteers treated with colchicine had a lower sialic acid content and a decreased aggregation potential in the presence of pathogenic LCs. These findings suggest that colchicine may be useful in the treatment of cast nephropathy and that it is conceivable to design peptides or analogs that would inhibit interactions of LCs with THP and theoretically prevent myeloma CN.

Cast formation may not rely only on interactions between LC and THP. First, 5 of 12 LCs purified from the urine of patients with CN failed to react with THP.16 Second, myeloma casts occasionally do not stain for THP in human
biopsies, and casts induced in mice by LC injection do not seem to contain THP during the first 24 hours, indicating that some LCs may undergo aggregation or precipitation in the absence of THP. This hypothesis is supported by studies showing that the deposition of certain LCs in vivo may be related to their capability to aggregate in vitro.44 Resistance of LCs to renal and macrophage-released proteases may also contribute to cast formation and persistence.16

Role of Tubular Obstruction by Casts in the Genesis of Renal Failure. The role of casts as plugs obstructing the tubules has been clearly shown in micropuncture studies. In myeloma patients, the correlation between severity of renal insufficiency and the number of casts remains controversial.24,25,45 This may be explained partly by the prominent medullary localization of casts, the count of which is underestimated in superficial kidney cortex biopsy specimens. The first indication that antibodies to THP could serve as probes of tubular obstruction was provided by Cohen and Border,46 who identified the protein in glomerular urinary spaces of two myeloma patients. This finding is indicative of intratubular urinary backflow. We detected THP in glomerular urinary spaces in 16 of 18 biopsies of patients with myeloma CN (Ronco and Mougenot, personal data) (Fig. 60.2). The proportion of obstructed tubules is too small to account by itself for renal failure. Renal failure induced by CN is multifactorial, implicating also tubular epithelial cell and interstitial lesions. Tubule obstruction by casts may explain the slow recovery of renal function noted in many patients.25

Interstitial deposits of THP were also found in 8 (44%) of the 18 biopsies (Ronco and Mougenot, personal data). They probably result from a leakage of the protein through gaps in the tubular basement membrane favored by tubular obstruction. Clinical and experimental models have implicated the protein in the pathogenesis of tubulointerstitial nephritis. Thomas et al.47 identified a single class of sialic acid-specific cell surface receptors for THP on polymorphonuclear leukocytes, and further showed that in vitro activation of human mononuclear phagocytes by particulate THP led to the release of gelatinase and reactive oxygen metabolites, both probably contributing to tissue damage.






FIGURE 60.2 Myeloma cast nephropathy. Immunofluorescence stain with anti-Tamm-Horsfall protein (THP) monoclonal antibody. Glomerular deposits in Bowman’s space delineate the inner aspect of Bowman’s capsule and penetrate between lobules of the capillary tuft. Identification of THP in the urinary spaces of glomeruli supports the obstructive role of casts with reflux of tubular urine. (Magnification, × 312.)


Clinical Presentation

Changing Presentation of Patients with Myeloma-Induced Renal Failure. When DeFronzo et al. reported the first series of 14 myeloma patients with acute renal failure in 1960,48 it was established that renal failure occurred at some time during the illness in approximately half of the patients, but that the mode of presentation was usually chronic with a slow progression over a period of several months to years.

The mode of presentation of renal failure in myeloma has changed dramatically over the years. In their review of 141 patients treated in Nottingham between 1960 and 1988, Rayner et al.49 showed that the absence of severe renal impairment at presentation predicted a low probability of developing renal failure subsequently. In only 5 of 34 patients of our own renal series25 did the diagnosis of myeloma antedate the discovery of renal failure by more than 1 month. In three patients, the presence of a monoclonal Ig was known for 10 to 18 years, but it only showed criteria of malignancy for less than 9 months. Two-thirds of the 107 patients referred to the Oxford Kidney Unit from 1987 to 200650 had myeloma diagnosed after their admission with acute kidney injury (AKI). More aggressive treatment of myeloma and higher awareness of the conditions that induce CN in the last two decades may have prevented LC precipitation within the tubule lumen in the patients with an established diagnosis of myeloma.

Demographic and Hematologic Characteristics of Patients with Cast Nephropathy-Related Renal Failure. Table 60.2 summarizes the clinical and pathologic data in four large series of myeloma patients with acute renal failure in which a renal biopsy was performed in at least 40% of the patients. A diagnosis of myeloma CN was established histologically in 81 of 99 (82%) renal biopsies, and lesions compatible with this diagnosis were found in 10 further biopsy specimens (10%). In comparison with the Mayo Clinic series of 869 unselected myeloma cases52 in which the mean age was 62 years and the male-female ratio was 1.55, patients with acute renal failure did not show any demographic particularity. Myeloma patients with renal failure are characterized by high tumor mass and virtually constant urinary LC loss, often of high output.

More than 70% of patients in the renal series have a high tumor burden (Table 60.2). This is confirmed by the
Alexanian series, which included 494 consecutive patients referred to an oncology center (Table 60.3).53 Only 3% of patients with myeloma of low tumor mass had renal failure, whereas 40% of those with high tumor burden had a serum creatinine greater than 180 µmol/L. These data contrast with the hematologic characteristics of patients with other renal complications of dysproteinemia including FS, amyloidosis, and MIDD, in whom the monoclonal B lymphocyte or plasma cell proliferation is either malignant but usually of low magnitude, or often benign from a hematologic point of view.








TABLE 60.2 Clinical and Pathologic Characteristics of Patients with Myeloma-Induced Renal Failure of Presumed or Established Tubulointerstitial Origin



























































Series


No. of Patients


Age (yr)


Male-Female Ratio


Tumor Mas


Serum Creatinine (µmol/L)


Urinary Light Chain >2 g/day


Renal Lesions in Biopsy Specimen


IIB


IIIB


Rota et al.25


34


66 (33-90)


0.88


15%


73%


960 (164-2000)


53%


26 MCN
2 ATN
2 CIN


Pozzi et al.26


50


63 (47-60)


1.38


12%


82%


798 (273-1518)


41%a


16 “Myeloma kidney”b 8 other


Pasquali et al.27


25


60 (48-74)


2.12


24%


72%


891 (455-1391)


72%


25 MCN


Irish et al.51


56


67 (42-82)


1.33


22%


78%


811 (302-2600)


NA


16 MCN, 5 AINc


a Total proteinuria, including light chains.

b Presumably myeloma cast nephropathy.

c “Compatible with myeloma.”


AIN, acute interstitial nephritis; ATN, acute tubular necrosis; CIN, chronic interstitial nephritis; MCN, myeloma cast nephropathy; NA, not available; IIB, intermediate tumor mass; IIIB, high tumor mass.









TABLE 60.3 Relation Between Tumor Mass and Renal Function

































Tumor Mass


No. of Patientsa


% of Patients with Serum Creatinine (µmol/L)


<180


180-270


>270


Low


151


97


1


2


Intermediate


183


89


5


6


High


160


60


17


23


a This series included 494 consecutive, previously untreated patients with multiple myeloma.


From Alexanian R, Barlogie B, Dixon D. Renal failure in multiple myeloma: pathogenesis and prognostic implications. Arch Intern Med. 1990;150:1693, with permission.


Another salient feature of myeloma associated with renal failure is the high prevalence of pure LC myelomas. Although they represent only about 20% of all myelomas, they are found in between 37% and 64% of patients with renal failure of presumed or established tubulointerstitial origin. Development of CN in two studies in which this diagnosis was established histologically25,27 was associated with urinary excretion of LCs exceeding 2 g per day in 53% and 72% of the patients (Table 60.2). LC protein excretion emerges as a highly significant independent factor of renal failure on multivariate analysis (Table 60.4). The risk of developing renal failure is twice as high in patients with pure LC myeloma, and five to six times greater in patients with LC proteinuria greater than 2.0 g per day compared to those with proteinuria less than 0.05 g per day. This indicates that in patients producing complete Ig molecules, CN essentially occurs in those synthetizing an excess of LCs. The frequency of renal failure is identical in patients excreting κ or λ λ LCs. IgD myeloma has the greatest potential for causing renal disease.50 Hypercalcemia also is a prominent independent pathogenetic factor on multivariate analysis, with a risk of renal failure five times greater in those patients with corrected calcium greater than 2.87 mmol/L.53









TABLE 60.4 Features Associated with Renal Failure






































































No. of Patientsa


% with Renal Failure


p


All patients


494


18



Urinary LC (g/day)



>2.0


123


39


0.00001



0.05-2.00


149


17




<0.05


222


7



Myeloma protein type



Only LC protein


93


31


0.0003



Other


401


15



Serum calcium (mmol/L)b



>2.87


104


49


0.00001



≥2.87


390


10



a Same series of patients as in Table 60.3.

b Corrected calcium (mmol/L).


LC, light chain.


From Alexanian R, Barlogie B, Dixon D. Renal failure in multiple myeloma: pathogenesis and prognostic implications. Arch Intern Med. 1990;150:1693, with permission.


The Clinical and Urinary Syndrome of Myeloma Cast Nephropathy. CN-induced renal failure is remarkably silent. Clinical signs are due to myeloma (or to hypercalcemia), including weakness, weight loss, bone pain, and infection. Because of their nonspecificity and their frequency in older patients, they often do not lead patients to take medical advice or physicians to prescribe serum and urinary electrophoreses, which are the key laboratory investigations for the diagnosis of myeloma. Peaks visible on serum or urine electrophoresis are then identified by immunoelectrophoresis or immunofixation. A preserved corrected calcium at presentation in patients with unexplained renal failure should alert clinicians to the possibility of myeloma.








TABLE 60.5 Precipitants of Acute Renal Failure in Myeloma



























































Series


No. of Patients


Dehydration


Sepsis


Hypercalcemia


Contrast Medium


NSAIDs


None


Rota et al.25


34a


65%


44%


44% (>2.60 mmol/L)


0%


24%



Pozzi et al.26


50a


24%


10%


34% (≥2.60 mmol/L)


4%


0%


44%


Ganeval et al.28


80b


10%


9%


30%


11%



35%


Irish et al.51


56a


4%


4%


23%


0%


11%


57%


Haynes et al.50


107


6%


5%


16% (> 2.90 mmol/L)



18%


65%


a Renal lesions are described in Table 60.2.

b Includes 19 patients with myeloma cast nephropathy, two with amyloidosis, and eight with LCDD (light- and heavy-chain deposition disease). NSAIDs, nonsteroidal anti-inflammatory drugs.


The main urinary feature is the excretion of a monoclonal LC, which accounts for 70% or more of total proteinuria in 80% of patients.25 LC proteinuria is usually not detected by urinary dipsticks, but only by techniques measuring total proteinuria. Certain LCs fail to react or react weakly in some widely used precipitation assays, such as the sulfosalicylic acid method, leading to falsely negative or underestimated results. The remaining proteins are composed of albumin and low molecular weight globulins that have failed to be reabsorbed by proximal tubule cells. In the rare patients with albuminuria greater than 1 g per day, CN is usually associated with glomerular lesions due to amyloidosis or MIDD. There is no hematuria in pure CN.

Precipitants of Cast Nephropathy. These are of paramount importance because of measures to prevent precipitation (Table 60.5). It is often difficult to identify a particular event responsible for precipitating renal failure, as these patients experience many of the complications of the disease at once, a common thread of which seems to be an effect on renal perfusion.

Hypercalcemia is an important precipitant found in 16% to 44% of the renal series (Table 60.5), and in 57% of the patients with renal failure in Alexanian nonrenal series.53 Presumably, hypercalcemia acts by inducing dehydration as a result of emesis and a nephrogenic diabetes insipidus. It may also enhance LC toxicity and cause nephrocalcinosis.


Dehydration, with or without hypercalcemia, and infection are other major risk factors for acute renal failure. Rota et al.25 found a high rate of urinary infections (10/34, 29%), which were associated in three cases with an increased proportion of polymorphonuclear leukocytes in the renal biopsy, suggesting an etiologic link between infection and deterioration of renal function. Infection also operates by causing dehydration and prompting the use of nephrotoxic antibiotics.

Contrast media have hitherto been considered an important precipitant of acute renal failure. It was hypothesized that the contrast medium bound to intratubular proteins, especially the LC and THP, causing them to precipitate and obstruct tubular flow. Contrast media also have vasoconstrictive effects, decreasing GFR and urinary output. McCarthy and Becker54 reviewed seven retrospective studies of myeloma patients receiving contrast media, involving 476 patients who had undergone a total of 568 examinations. The prevalence of acute renal failure (which was not defined) was 0.6% to 1.25%, compared to 0.15% in the general population. This is a low risk and contradicts the dogma that contrast media should not be used in myeloma patients. This change may reflect awareness of the risk and care taken to hydrate patients actively with alkaline solutes before and during the administration of contrast media. No clinical data currently support the preferential use of nonionic agents in myeloma patients to decrease the risk of acute renal failure.

A number of drugs are noxious in myeloma patients. They include antibiotics, particularly aminoglycosides, and nonsteroidal anti-inflammatory drugs (NSAIDs).25,50 NSAIDs reduce the production of vasodilatory prostaglandins that help to maintain an appropriate GFR in patients with renal hemodynamics compromised by dehydration. Angiotensinconverting enzyme (ACE) inhibitors can also precipitate renal failure because they reduce GFR dramatically in dehydrated patients. Their use as that of angiotensin type-1 receptor antagonists should be avoided as long as a risk of decreased renal perfusion persists.

Recently introduced therapies including bisphosphonates may also induce toxic tubular injury. Renal failure secondary to acute tubular necrosis was reported with zoledronate, a potent bisphosphonate that is in widespread use for the treatment of hypercalcemia of malignancy,55 and with short-term, high-dose pamidronate.56 Doses should be adapted to GFR to avoid toxicity.


Renal Pathology and the Value of Kidney Biopsy

A kidney biopsy should not be routinely performed in patients with a presumed diagnosis of myeloma CN. However, it is useful in three circumstances:

1. To establish the cause of renal failure in anuric patients with clinically silent myeloma without evidence of serum monoclonal component on electrophoresis;

2. To analyze tubulointerstitial lesions and predict the reversibility of renal failure in patients with presumed CN but multiple precipitating factors;

3. To identify glomerular lesions in patients with urinary albumin greater than 1 g per day and no evidence of amyloid deposits in “peripheral” biopsies (accessory salivary glands, rectum, abdominal fat).

A kidney biopsy should be systematically performed in patients enrolled in therapeutic protocols because renal lesions should be precisely identified.57

Myeloma Casts. Myeloma CN is characterized by the presence of specific casts associated with severe alterations of the tubule epithelium. Myeloma casts are large and usually numerous. Their prevailing localization is the distal tubule and the collecting duct, but they may also be found in the proximal tubule and even in the glomerular urinary space. They often have a “hard” and “fractured” appearance, and show polychromatism upon staining with Masson’s trichrome (Fig. 60.3). Casts may also have a stratified or laminated appearance. They may stain with Congo red, but only exceptionally do they show the typical yellow-green dichroism of amyloid under polarized light.

An important diagnostic feature of myeloma casts is the presence of crystals, which may be suspected by light microscopy.58 Such casts are often angular or heterogeneous because they contain multiple rhomboid or needle-shaped crystals surrounded by amorphous material and cell debris.

Casts are frequently surrounded by mononuclear cells, exfoliated tubular cells, and, more characteristically, by multinucleated giant cells whose macrophagic origin has been established by specific antibodies. These cells are often seen engulfing the casts and at times actually phagocytizing fragments. In some cases, the cellular reaction is made
of polymorphonuclear leukocytes in the absence of urinary tract infection. Typical myeloma casts with a giant, multinucleated cell reaction (Fig. 60.3) can be very occasionally detected in other hemopathies including µ-HC disease30 and Waldenstöm’s macroglobulinemia.29 In myeloma CN, there is a great variability in the respective percentage of typical myeloma casts and of nonspecific hyaline casts. In some instances, most casts have nonspecific characteristics by light microscopy, even if by immunofluorescence the vast majority consists predominantly of one of the two LC types. The search for typical casts has to be conducted on all available sections if necessary.






FIGURE 60.3 Myeloma cast nephropathy. Typical myeloma casts with fractured appearance are surrounded by multinucleated macrophagic cells (arrows) in a patient with λ-light-chain myeloma. (Masson’s trichrome, ×312.)

By immunofluorescence, myeloma casts are essentially composed of the monoclonal LC excreted by the patient, together with THP. In most cases, casts are stained exclusively or predominantly with either the anti-κ or the anti-κ antibody. However, in about 25% of myeloma biopsies, casts stain for both antibodies because they contain polyclonal LCs, together with albumin and fibrinogen.58 Staining of “angular” casts is often irregular, and more intense at the periphery (Fig. 60.4). In heterogeneous casts, the crystals themselves fail to stain, whereas the matrix of the cast and the surrounding cellular debris and amorphous material often stain positively for one of the LC isotypes.

Cast ultrastructure was studied by electron microscopy in 24 biopsies of myeloma CN by Pirani et al.58 Crystals were detected in 14 biopsy specimens and suspected in another 4. The authors have identified four major categories of casts, according to their content and ultrastructural appearance. One category characterized by large rectangular crystals, or fragments thereof, with a pentagonal or hexagonal cross-section, is found only in myeloma CN. It seems to be closely linked to the development of a giant cell reaction around the cast. A second category also frequently contains crystals, but they are small, electron-dense, and needle-shaped, and seemingly not associated with a cellular reaction. Similar large rectangular and small, needle-shaped crystals can be found within plasma cells. They are also seen occasionally within the cytoplasm of either proximal or distal tubular cells (Fig. 60.5), surrounded by a single smooth membrane, which suggests that they are located within lysosomes.






FIGURE 60.4 Myeloma cast nephropathy. Several tubules contain large casts, one of which has an angular and fractured aspect. The stain with anti-κ antibody is more intense at the periphery of most casts. (Immunofluorescence, × 312.)






FIGURE 60.5 Myeloma cast nephropathy. Rectangular crystals presumably composed of λ-light chains in tubular cells.

(Electron micrograph, uranyl acetate, and lead citrate, ×7,000.)

Tubules and Interstitium. Considerable tubular damage is almost always present in myeloma CN. Epithelial tubular lesions are not only seen in the distal tubules where casts are principally located, but also in proximal convoluted tubules, where the epithelium undergoes atrophy and degenerative changes. Frank tubular necrosis may also be seen, with or without typical myeloma CN.25 By immunofluorescence, a variable number of tubule sections contain numerous “protein reabsorption droplets” staining for the monoclonal LC.46

Interstitial lesions are often associated with the tubular damage. They may be mild and consist of inflammatory infiltrates and fibroedema, but fibrosis and its correlate, tubular atrophy, may also be fairly extensive. In severe cases with epithelial denudation and gaps in the continuity of the tubular basement membrane, often in close contact with myeloma casts, granulomatouslike formations containing macrophages and histiocytes develop around the ruptured tubules (Fig. 60.6).46

Glomeruli and Vessels. The glomeruli are usually normal, except for small clusters of globally sclerotic glomeruli and a mild thickening of the mesangial matrix. When mesangial thickening is more prominent, the possibility of an associated
MIDD should be considered. Rarely, amorphous deposits reminiscent of myeloma casts can be seen in capillary loops or in the glomerular urinary space. In younger patients, severe chronic vascular lesions are sometimes observed, which may contribute to progression of sclerosis.






FIGURE 60.6 Myeloma cast nephropathy. Interstitial granulomatous-like formations with macrophages surrounding disrupted tubular basement membrane ( arrow ) were numerous in this λ-chain cast nephropathy. (Silver stain, ×312.)


Outcome and Prognosis of Myeloma Cast Nephropathy

Until the 1980s, myeloma-induced renal failure was associated with a very poor prognosis, with a median survival of less than 1 year.48 In recent years, the outcome of patients with myeloma, including those with renal impairment, has improved with the introduction of novel therapies, including high dose therapy followed by autologous stem cell support, and development of new drugs with a strong anti-myeloma effect (bortezomib, thalidomide, lenalidomide).59 However, because patients with elevated serum creatinine levels were excluded from most randomized controlled studies, optimal treatment of multiple myeloma with renal failure remains to be defined. The establishment of consensus criteria for the assessment of renal function and renal response in multiple myeloma, and the use of modern sensitive tests to evaluate hematologic response, such as as nephelometric assays for serum free LC,60 should help to improve renal and patient outcomes in the future.61

Renal Outcome and Prognostic Factors. Renal prognosis in patients with myeloma who present with renal failure remains poor, as complete or partial renal recovery occurs in half of patients after weeks to months,62 and in only 20% to 40% of those with dialysis-dependent renal failure.63,64,65,66 Elevated plasma creatinine concentration or decreased estimated GFR (calculated using the MDRD equation) have been quoted as markers of poor renal prognosis in most studies,26,28,62,65,67 implying that renal functional impairment of any degree should be treated as a medical emergency. In the study by Rota et al.,25 main prognostic indicators were provided by renal histology. Renal response was seen in patients with typical cast nephropathy and/or tubular necrosis without interstitial damage. Global tubular atrophy and interstitial fibrosis were associated with partially or totally irreversible renal failure, whereas the number of casts has a controversial predictive value.25,26,68

The rapid achievement of sustained hematologic response appears as a key factor for renal prognosis.57,65,69,70 In three recent studies a minimum of 50% reduction in serum free LC concentration was required for recovery of renal function in patients with biopsy proven CN.57,70,71 Recent data indicate that outcome of severe renal failure may be substantially improved by bortezomib plus dexamethasone-based chemotherapy61,65,66,72 and, in patients requiring dialysis, by extended hemodialysis using a high-cut off dialyzer that allows effective removal of LCs.70,73

Survival and Predictors. Myeloma patients with renal failure have a shorter survival than those with normal renal function. In the presence of renal failure, mortality in the first 3 months is about 30%,26,28,67 and median survival ranges from 9 to 22 months. However, several studies have indicated that recovery of renal function is associated with improved survival, close to that of patients who do not develop renal failure.25,26,27,28,50,67,62,69 A response to chemotherapy is a key predictor of renal outcome and patient survival.



Fanconi Syndrome

Fanconi syndrome (FS) is characterized by renal glycosuria, generalized aminoaciduria, hypophosphatemia, and, frequently, by chronic acidosis, hypouricemia, and hypokalemia. It often includes osteomalacia, with pseudofractures. These manifestations result from functional impairment of the renal proximal tubule. The first association of FS with myeloma was reported by Sirota and Hamerman,99 although these authors considered FS and myeloma as two separate diseases. Engle and Wallis100 identified crystal-like inclusions in both tumor cells and renal tubule epithelial cells, and suggested that FS and myeloma could be related. Costanza and Smoller101 described the cytoplasmic inclusions as round or rodlike electron-opaque structures with longitudinally oriented fibrils. Lee et al.102 established clearly that mye loma was a cause of adult FS. Maldonado et al.103 reported 17 cases of FS associated with plasma cell dyscrasia, and two more recent studies described the clinicopathologic features of the disease in two series of 11 and 32 patients.104,105 The disease is most likely underdiagnosed. The rarity of FS in patients with myeloma contrasts, however, with the high prevalence of tubule alterations in myeloma autopsy series. This suggests that unusual specific properties of LCs, mostly κ, are involved in the pathophysiology of FS.


Pathophysiology of Plasma Cell Dyscrasia-Associated Fanconi Syndrome

The peculiar propensity of certain LCs to form crystals in vivo is attested by experimental studies in mice14 and rats.40 It is remarkable that the κLCs that induced crystallization in vivo, also significantly reduced the glucose, chloride, and volume fluxes.

Crystal composition was analyzed in a patient with myeloma-associated FS and hexagonal crystals in kidney proximal tubular cells, bone marrow plasma cells, and phagocytes.106 N-terminal sequencing and mass spectrometry studies showed that a 107-amino acid fragment corresponding to the variable domain of the κ-LC (V#c) was the essential component of crystals forming spontaneously from the patient’s urine (Fig. 60.7A). Vκ was also crystallized alone using the hanging drop technique (Fig. 60.7B). Crystals were hexagonal bipyramids and had the same 6.0-nm periodicity on electron micrographs as those found in the cells. The V domain (12-kDa) resisted proteolysis by trypsin, pepsin, and cathepsin B, self-reacted, and formed crystals in vitro, which may explain its accumulation in plasma cells and proximal tubular cells. The resistance of LC V domains to proteolytic enzymes including cathepsin B was confirmed in further studies,16,104 except in a few patients with a high-mass myeloma or Waldenström’s macroglobulinaemia107 and FS. At variance with the observations made in patients with CN,16 LCs from patients with FS did not bind THP, except in one case where both syndromes were associated.

The unusual physicochemical behavior of FS κ-chains was tentatively correlated with their structure in a number of cases.108 Sequence analyses showed that 90% of LCs belonged to the VλVI variability subgroup, whereas this subgroup only accounts for 56% of all monoclonal κ-LCs.104,109 The VλVI appeared to originate from only two germline genes, IGKV1-39 in five cases and IGK1-33 in four. Analyses of the DNA sequence suggested that all structure peculiarities arose from somatic mutations in the proliferating clone.110 In the 10 available sequences, residues had never or rarely been reported among VλVI subgroup LCs. The unusual presence of nonpolar or hydrophobic amino acids in the complementary determining region (CDR)-L1
loop at position 30, together with a nonpolar amino acid at position 50, seems to be specific for FS LCs derived from gene IGKV1-39. These hydrophobic residues are exposed to the LC surface108 and may be involved in the pathophysiology of FS, as is suggested by site-directed mutagenesis in an experimental model for FS.111 Recently, a transgenic mouse model with overexpression of human FS κ LC (CHEB) was generated through insertion of the V domain from CHEB LC in the Igκ locus of the mouse. This resulted in the expression of a hybrid κ LC made up of the human V domain and the mouse constant (C) region. Despite the replacement of the human C region, animals exhibited characteristic LC crystals within proximal tubular cells. This model confirmed that LC aggregation in FS is promoted by the V domain structure. The extent of tubular inclusions was proportional to the production rate and serum levels of CHEB LC. Using an inducible CRE mediated deletion of the CHEB V domain, tubular lesions were shown to recover after a few weeks when the LC production was stopped.112






FIGURE 60.7 Plasma cell dyscrasia-associated Fanconi syndrome. Crystals spontaneously obtained in vitro from a Sephadex G100 fraction of the patient’s urinary proteins (A) and by the hanging drop technique from purified Vκ fragment (B). (A, magnification, ×400; B , size of these crystals, 0.25 mm.) (From Aucouturier P, Bauwens M, Khamlichi AA, et al. Monoclonal Ig L chain and L chain V domain fragment crystallization in myeloma-associated Fanconi’s syndrome. J Immunol. 1993;150:3561, with permission.)

After endocytosis, LCs are processed in the endosomal and lysosomal compartment where “normal” LCs are degraded. In FS, accumulation of the protease-resistant V domain fragment generated by lysosomal enzymes may induce crystal formation. Clogging of the endolysosomal system may subsequently alter apical membrane recycling and/or adenosine triphosphate (ATP) production (hence, Na++-ATPase functioning) as suggested by mitochondrial injury,102 and lead to progressive impairment of sodium-dependent apical transporters. However, in a few cases of FS, crystalline inclusions were not observed within proximal tubular cells.104,107,108 In two patients, LCs, which belonged to the VkIII subgroup, showed no common substitution with previously described FS VκI LC and did not display resistance to proteolysis, suggesting that other mechanisms of toxicity may be involved in the pathogenesis of the disease.10 Furthermore, why FS does not occur in patients with apparently the same degree of distortion of the lysosomal compartment as can be seen in certain myeloma patients with or without CN is unclear. The molecular mechanisms responsible for glycosuria, phosphaturia, generalized aminoaciduria, and uric acid loss remain poorly understood. An impairment of the megalin-cubilin system might be involved.


Clinical Presentation

The clinical features are summarized in Table 60.6.104 The median age at diagnosis is 57 years. Most common initial manifestations are bone pain and weakness, principally due to osteomalacia. The major cause of this osteomalacia is hypophosphatemia, which results from increased urinary clearance of phosphate. Chronic acidosis and abnormal renal vitamin D metabolism further contribute to the development of bone lesions. Bone pain may also be the consequence of lytic lesions in patients with a high-mass myeloma. Other revealing signs are essentially due to the proximal tubule impairment, including hypokalemia. Renal failure occurs more frequently than one would expect in a disease of the proximal tubule.

Criteria for the diagnosis of FS may not all be present together, especially in patients with renal failure.109 The diagnosis of FS is often unrecognized for several years in patients presenting with proteinuria, bone pain, or renal failure. The mean time from onset to diagnosis of FS is about 3 years.104 Typically, the diagnosis of FS precedes that of the plasma cell dyscrasia, most often a κ-LC-excreting multiple
myeloma, because the hematologic disease has a low tumor burden and a slow progression. In 35 of 98 (36%) published cases,104,105 even criteria for the diagnosis of myeloma were lacking, and patients were classified initially as having a benign monoclonal gammopathy of undetermined significance (MGUS). In some patients, the diagnosis of the plasma cell dyscrasia remained undetermined between myeloma and MGUS because it may be difficult to recognize the cytologic characteristics of myeloma cells when their cytoplasm is stuffed with crystals. Three patients of 99 had Waldenström’s macroglobulinemia.104,105,107








TABLE 60.6 Clinical Characteristics of Patients with Plasma Cell Dyscrasia-Associated Fanconi Syndromea

























Total No. of Patients


Age Mean/Extremes


Gender


Initial Manifestations


Bone Lesions


Renal Failurec


Plasma Cell Dyscrasia


Light-Chain Isotype


68


57 22-81


30 males 38 females


Bone pain (25)b Weakness, fatigue (16) Weight loss (7) Polyuria-polydipsia(7) Hypokalemiarelated signs (4) Proteinuria (18) Renal failure (16) Renal glycosuria (13)


Osteomalacia (25) High-mass myeloma (12) Plasmacytoma (1)


54


Myeloma (36)d MGUS (21)e MGUS/myeloma (4)f Lymphoma/CLL (4)g “Atypical” plasma cell dyscrasia (1)


49κ



aFigures in parentheses indicate number of patients.

b Related to osteomalacia.

c Serum creatinine >130 µmol/L, or creatinine clearance

d Including 12 patients with a high-mass myeloma.

e Monoclonal gammopathy of undetermined significance (MGUS).

f Undetermined diagnosis, mostly due to cytoplasmic inclusions in plasma cells making interpretation of cytology difficult.

g Chronic lymphocytic leukemia (CLL).


From Messiaen T, Deret S, Mougenot B, et al. Adult Fanconi’s syndrome secondary to light-chain gammopathy: clinicopathologic heterogeneity and unusual features in 11 patients. Medicine (Baltimore). 2000;79:135, with permission.


Conversely, the metabolic disorders of FS may be overseen in the context of myeloma, and FS-related bone lesions should not be interpreted as the consequence of high-mass myeloma.


Pathologic Data

Typically there are prominent crystals in enlarged proximal tubular cells and degenerative changes of proximal tubules.104 Proximal tubular cells are stuffed with microcrystals that stain red or green with Masson’s trichrome and are periodic acid-Schiff negative. In the most severely affected tubules, crystal-containing exfoliated cells are seen in the tubular lumen, whereas intracytoplasmic crystals are still present in atrophic tubules. In other cases, crystals can only be suspected by the presence of a finely granular material of glassy appearance in an enlarged proximal tubular epithelium (Fig. 60.8A). Their presence is more easily demonstrated by toluidine-blue staining of semi-thin sections and by hematoxylin and eosin staining of cryostat sections. In the same tubule sections, all the cells are not equally affected; cells with a normal aspect coexist with those stuffed with crystals.

A universal feature is the additional presence of severe lesions of the proximal tubule epithelium apparently devoid of crystals. These lesions include vacuolization, loss of the luminal brush border, and focal cell sloughing, with cell fragments in the lumen of the tubules. Interstitial cellular infiltrate, including plasma cells, may contain crystalline inclusion bodies. Patchy tubular atrophy and focal interstitial fibrosis, together with a variable number of obsolescent glomeruli, are often observed.

In several cases, attempts to characterize the crystal proteins with anti-Ig conjugates, including anti-LC antibodies, have failed. When immunohistochemical studies are positive, crystals stain only (or predominantly) for the monoclonal LC, most often κ (Fig. 60.8B).

By electron microscopy, crystals of various size and shape (rectangular, rhomboid, round, or needle-shaped) are detected within the cytoplasm of proximal tubule cells
(Fig. 60.8C). Intracytoplasmic crystals are surrounded by a single smooth membrane, likely of lysosomal origin.58,106 In rare cases, crystals are also seen in distal tubule cells. In other cases, crystals are not found by light microscopy, but electron microscopy shows enlarged vesicular bodies containing dense tubular and rod-like structures101,102,103 or fibrils and needle-shaped deposits very close to crystalline structures.






FIGURE 60.8 Plasma cell dyscrasia-associated Fanconi syndrome. A,B: Renal biopsy. A: Glassy appearance of the epithelium of several proximal convoluted tubules. Crystals were not evident by light microscopy, but were demonstrated by electron microscopy. Note also the severe lesions of the epithelial cells lining some tubules (arrow), and mild interstitial fibrosis. (Masson’s trichrome, ×312.) B: Immunofluorescence stain of the same tubules with anti-κ monoclonal antibody. (Magnification, ×312.) C: Bone marrow smear from the same patient. The cytoplasm of a plasma cell shows a vacuolated aspect, suggesting the presence of crystals. (May-Grünwald-Giemsa stain, ×1,000.)

Crystal formation in plasma cell dyscrasia-associated FS is not limited to renal tubule epithelium but also occurs in bone marrow and tissue-infiltrating plasma cells, and in macrophages (Figs. 60.8C and 60.9).103,106 In plasma cells, crystals are localized not only in lysosomes, but they are also frequently found inside the granular endoplasmic reticulum. Crystal formation in these organelles therefore suggests incomplete proteolysis of LCs. The slow progression of myeloma disease in typical FS associated with crystal formation may be explained by the deleterious effects on cell growth of the accumulation of crystalline inclusions in the tumor plasma cells. A peculiar accumulation of LC crystals may be observed within lysosomes of macrophages in the bone marrow and other organs, defining “crystal-storing histiocytosis” (CSH). In CSH, invariably associated with κ LC monoclonal gammopathy, crystals appear to be mostly made up of monoclonal κ LC, and more rarely of entire IgG. Renal manifestations in CSH are mostly represented by chronic tubulointerstitial nephritis and FS with accumulation of monoclonal κ LC crystals within proximal tubular cells. Perirenal and interstitial infiltration by histiocytes containing eosinophilic crystalline inclusions (pseudo- pseudo Gaucher cells) is suggestive of the disease. Specific molecular peculiarities in the V domains of CSH monoclonal κ LCs may account for crystal accumulation within histiocytes and multiple organ involvement.113

Although crystals are a salient feature of the plasma cell dyscrasia-associated FS, they are neither specific nor absolutely constant. Crystals were found in 16 of 28 (57%) patients in the two largest series published so far.104,105 They may also be found, albeit in low amounts, in proximal tubule epithelial cells of patients with CN,16,58 and occasionally in myeloma patients with isolated tubular lesions, that is, in the absence of myeloma casts.


Outcome and Treatment

As expected, patients with multiple myeloma have shorter survival time than those with MGUS. In the Mayo Clinic’s series, only one of the 14 patients with MGUS developed multiple myeloma, and at the end of follow-up, only 5 of 32 patients had evolved to end-stage renal disease.105

In patients with osteomalacia, considerable improvement can be obtained with 1α-hydroxyvitamin D, calcium, and phosphorus supplementation. The effect of chemotherapy on the proximal tubule impairment is much more debated. It was reported that the treatment of underlying myeloma improved urinary signs and tubular transport abnormalities. However, no significant change in renal function was observed in the two largest series.104,105 It has been suggested that the presence of crystals within plasma cells should be added to the list of criteria against chemotherapy in
myeloma. Because chemotherapy, especially with alkylating agents, carries a significant risk of complications but without much benefit for kidney function, and because patients who do not have an overt malignancy show a relatively benign course, the risks and benefits of chemotherapy should be weighed carefully. Whether novel antimyeloma agents might improve renal prognosis in FS remains to be established.






FIGURE 60.9 Plasma cell dyscrasia-associated Fanconi syndrome in same patient as in Figure 60.7. Electron microscopic study of intracellular (A, B, and C) and in vitro-formed crystals in the same patient. A: Bone marrow plasma cell (and a macrophage on the left). (Magnification, ×8,000.) B: Bone marrow macrophage. (Magnification, ×50,000.) C: Proximal convoluted tubular epithelial cell. (Magnification, ×50,000.) D: Crystal obtained in vitro from Sephadex G100 fraction C from the patient’s urine. (From Aucouturier P, Bauwens M, Khamlichi AA, et al. Monoclonal Ig L chain and L chain V domain fragment crystallization in myeloma-associated Fanconi’s syndrome. J Immunol. 1993;150:3561, with permission.)


AMYLOIDOSIS

Amyloidosis has been known to be associated with or to cause renal disease for more than 100 years. Amyloid was originally identified as a waxy substance by Rokitansky in 1842, but the term amyloid was coined by Virchow in 1854 because the substance stained with iodine in a way that was similar to starch and cellulose. Although the protein content of amyloid was recognized subsequently, the term amyloid persisted. The diversity of amyloidotic disease was rapidly suspected on clinical grounds, but chemical studies in the late 1960s actually provided the basis of the present classification of amyloid (Table 60.7). In 1968, Pras et al.114 isolated and purified amyloid fibrils, which opened the way to further chemical analyses. In 1971, Glenner et al.2 found that the amyloid fibril proteins from two patients had an N-terminal sequence identical to Ig LCs, which was the first demonstration of a relation between amyloidosis and Ig. They also generated “amyloidlike” fibrils by proteolytic digestion of some human LCs, thereby demonstrating their propensity for forming amyloid.115

AL amyloidosis is certainly among the most severe complications of plasma cell proliferative disorders. The only efficient therapeutic tools to date are chemotherapeutic drugs against B cell proliferations. However, pathophysiologic considerations and advances in the treatment of other types of amyloidosis may open new therapeutic avenues (Table 60.7).


General Characteristics of Amyloidosis


A Common Ultrastructural Molecular Organization Defining a Morphologic Entity

Amyloidosis is the general term for a morphologic entity, defined by visceral, extracellular deposition of protein material with unique tinctorial properties and ultrastructural characteristics. After Congo red staining, amyloid deposits exhibit
birefringence under polarized light, which indicates the presence of highly ordered structures. These deposits have been extensively studied at the ultrastructural level by electron microscopy, infrared spectroscopy, and X-ray diffraction. Glenner116 clustered all amyloidoses under the denomination of β-fibrilloses on the basis of the highly similar organization of the amyloid deposits. These are “typically composed of a felt-like array of 7.5- to 10-nm wide rigid, linear, nonbranching, aggregated fibrils of indefinite length.” One amyloid fibril is made of two twisted 3-nm-wide filaments, each having a regular antiparallel β-pleated sheet configuration; the β-sheets are perpendicular to the filament axis. A regular packing of peptides or proteins with a β-sheet conformation results in the elongation of amyloid fibrils. The numerous hydrogen bonds between virtually all amide functions of the peptide backbones make such a structure highly stable. Other components, described in subsequent text, are supposed to stabilize the fibrils.








TABLE 60.7 Classification of Amyloidoses











































































































Amyloid Protein


Precursor


Distribution


Type


Syndrome or Main Involved Tissues


AA


Serum amyloid A


Systemic


Acquired


Secondary amyloidosis, reactive to chronic infection or inflammation including hereditary periodic fever (FMF, TRAPS, HIDS, FCU, and MWS)


AApoAI


Apolipoprotein A-I


Systemic


Hereditary


Liver, kidney, heart, skin, larynx


AApoAII


Apolipoprotein A-II


Systemic


Hereditary


Kidney, liver, adrenal glands, spleen, skin



Aβ protein precursor


Localized Localized


Acquired Hereditary


Sporadic Alzheimer disease, aging Prototypical hereditary cerebral amyloid angiopathy, Dutch type


Aβ2M


β2-microglobulin


Systemic


Acquired


Chronic hemodialysis


ABri


Abri protein precursor


Localized or systemic?


Hereditary


British familial dementia


ACys


Cystatin C


Systemic


Hereditary


Icelandic hereditary cerebral amyloid angiopathy


AFib


Fibrinogen Aα chain


Systemic


Hereditary


Kidney


AGel


Gelsolin


Systemic


Hereditary


Finnish hereditary amyloidosis


AH


Immunoglobulin heavy chain


Systemic or localized


Acquired


Primary amyloidosis, myeloma-associated


AL


Lysozyme


Systemic


Hereditary


Kidney, liver, spleen, adrenal glands


APrP


Prion protein


Localized


Acquired


Sporadic (iatrogenic CJD, new variant CJD) (alimentary?)




Localized


Hereditary


Familial CJD, GSSD, FFI


ATTR


Transthyretin


Systemic


Hereditary Acquired


Prototypical FAP Senile heart, vessels


ALECT2


Leukocyte chemotactic factor 2


Systemic


Acquired?


Kidneys, liver, adrenal glands


Lines in bold characters indicate amyloid types with kidney involvement.


The following proteins may also cause amyloidosis: calcitonin, islet-amyloid polypeptides, atrial natriuretic factor, prolactin, insulin, lactadherin, keratoepithelin, and Danish amyloid protein (which comes from the same gene as ABri and has an identical N-terminal sequence).


CJD, denotes Creutzfeldt-Jakob disease; FAP, familial amyloidotic polyneuropathy; FCU, familial cold urticaria; FFI, fatal familial insomnia; FMF, familial Mediterranean fever; GSSD, Gerstmann-Strüassler-Scheinker disease; HIDS, hyper-IgD syndrome; MWS, Muckle-Wells syndrome; TRAPS, tumor necrosis factor receptor-associated periodic syndrome.


Adapted from Westermark G, Benson MD, Buxbaum JN, et al. Amyloid fibril protein nomenclature—2002. Amyloid 2002;9:97; Merlini G, Bellotti V. Molecular mechanisms of amyloidosis. N Engl J Med. 2003;349:583.




Amyloid Protein Precursors and Classification of Amyloidoses

Amyloid protein precursors share the property of either a native β-pleated conformation or a high propensity to form β-sheets. All are globular structures, clearly distinct from fibrillar proteins such as collagen, which are proline-rich polymers with a longitudinal arrangement.






FIGURE 60.10 Amyloidosis. A: Glomerular and vascular heavy amyloid deposits stained with antitransthyretin antibody in a patient with Portuguese-type hereditary amyloidosis. (Immunofluorescence, ×312.) B: Co-deposition of amyloid P (AP) component in a glomerulus from the same patient as in (A). (Immunofluorescence stain with anti-AP component antibody, ×312.) C: Glomerulus with early amyloid deposits in mesangium, capillary walls, and arteriolar wall (arrows) from a patient with AA amyloidosis. (Light microscopy, periodic acid-Schiff, ×312.) D: Glomerulus from a patient with AL amyloidosis. Scanty glomerular deposits contrast with almost complete replacement of arterial walls by amyloid. (Immunofluorescence stain with anti-κ antibody, ×312.)

The International Committee for Amyloidosis recommended a nomenclature essentially based on the nature of amyloid proteins117 ; the abbreviated name of each amyloid protein is preceded by the letter A. The list provided in Table 60.7 is not exhaustive. Twenty-seven different amyloid protein precursors have been identified to date. It is worth noting that hereditary and secondary forms of the same disease exist and should be distinguished; for instance, normal transthyretin is responsible for senile systemic amyloidosis, whereas certain mutations are the cause of familial amyloidotic polyneuropathy (Fig. 60.10A). Multiple different factors, either intrinsic (structural) or external (concentration of the precursor proteins, tissue factors, etc.), may influence the pathogenicity of a variety of potentially amyloidogenic proteins.


Other Constituents of Amyloid

In addition to the unique “pseudocrystalline” stacking of β-sheets, a few structural features are shared by all types of amyloid, and might help the understanding of some aspects
of the pathophysiology. Glycosaminoglycans (GAGs) have been found tightly associated with all isolated amyloid fibrils. GAGs are polysaccharide chains made of repeating uronic acid-hexosamine units of several types and normally linked to a protein core, thus constituting proteoglycans, which are important constituents of extracellular matrices. The invariable presence of GAGs in amyloid fibrils raises two suggestions:

1. Proteoglycans might interact with amyloidogenic precursors during the nucleation steps of amyloidogenesis; indeed, most GAGs associated with fibrils are of the heparan sulfate type, and heparan sulfate proteoglycans are essential components of the basement membranes, which are preferential sites of amyloid deposition. Recent data indicate that specific interactions occur between motifs within heparan sulfate and properly modified AL LCs.118

2. Sulfated GAGs might be important for inducing and stabilizing the β-pleated structure of the amyloid fibrils.119

Another constituent of all amyloid deposits is a protein of the pentraxin family, the serum amyloid P component (SAP) (Fig. 60.10B). SAP is a plasma glycoprotein made up of two noncovalently linked pentamers of identical subunits. The β-pleated structure of SAP120 is strongly homologous to that of legume lectins such as concanavalin A. It shows no allelic polymorphism and displays striking interspecies homology. Furthermore, no occurrence of SAP deficiency has yet been described, which suggests that it has essential physiologic functions. SAP is a calcium-dependent lectin, with binding affinities toward DNA, C4-binding protein, and the collagenlike region of C1q, and several constituents of extracellular matrices such as fibronectin and proteoglycans. SAP was shown to bind apoptotic cells and nuclear debris, and mice with targeted deletion of the SAP genes spontaneously develop anti-DNA antibody and a syndrome resembling human systemic lupus erythematosus.121 Two calcium sites are involved in carbohydrate binding. In the presence of calcium, SAP is remarkably resistant to proteolytic digestion, suggesting a physiologic role in maintaining extracellular matrix structures. Coating of amyloid fibrils with unaltered SAP is a constant feature that could result in their protection from catabolism. It is probable that SAP binding to amyloid deposits is mediated by GAGs through the formation of multicomponent complexes. The high affinity of SAP toward all types of amyloid is used for diagnosing and monitoring the extent of systemic amyloidosis using scintigraphy with123 I-labeled SAP.122 SAP binding to all ligands is inhibited by specific sugars such as β – D – galactose cyclic pyruvate acetal. Moreover, the knowledge of SAP structure offers the opportunity of designing competitive inhibitors as potential drugs for the treatment of amyloidoses. Recently, CPHPC, a compound that specifically binds to SAP allowing a rapid decrease in serum SAP levels, has been developed.123 The combination of CPHPC with an antibody specific for SAP, which targets amyloid deposits and enables their elimination by recruiting phagocytic cells, has shown impressive results in an experimental mouse model of systemic AA amyloidosis.124


General Mechanisms of Fibrillogenesis

The amyloidoses are diseases of protein conformation in which a particular soluble innocuous protein transforms and aggregates into an insoluble fibrillar structure that deposits in extracellular spaces of certain tissues. Fibrillogenesis may be the consequence of several mechanisms of processing the amyloid precursor, including partial proteolysis and conformational modifications. In systemic AA amyloidosis, removing of the C-terminal part of an apolipoprotein acutephase reactant, SAA, yields a 5- to 10-kDa fibril-forming fragment. Phagocytic cells, in particular macrophages, supposedly play a central role in this disease by providing the intralysosomal processing of the precursor. In other forms of amyloidosis, such as those involving transthyretin and Ig LCs, partial proteolysis has been demonstrated but may as well occur after fibrillogenesis, as shown in AA amyloidosis. The demonstration of small fragments from the LC constant domain in deposited fibrils also argues in favor of a postfibrillogenic proteolysis in AL amyloidosis.

In certain types of hereditary amyloidoses due to mutations in the genes coding for the precursor protein,125 amyloid formation seems to occur via a conformational change leading to a soluble partially folded intermediate. The property shared by these amyloidogenic variants is a native conformation that is thermodynamically less stable than that of the normal counterpart. A reduction in the stability of the variant was shown to favor the formation of partially folded conformers (alternative spatial arrangements of the same polypeptide) that have a strong propensity to self-aggregate and assemble into fibrils. Whether conclusions from structural studies of transthyretin or lysozyme mutants may be extended to other amyloidoses, including AL and AA amyloidoses, remains questionable.

Amyloidogenesis seems to be a nucleation-dependent polymerization process. Unlike other protein deposition diseases, in which amorphous aggregates are the consequence of insolubility of the pathogenic protein in the tissues, amyloid may result from a “one-dimensional crystallization.” Formation of an ordered nucleus is the initial and limiting step, followed by a thermodynamically favorable addition of monomers leading to elongation of the fibrils. As shown in Alzheimer and prion diseases, the nucleation step can be overrun by adding a preformed nucleus to a supersaturated solution of the amyloidogenic protein. A similar “seeding” phenomenon may explain the “amyloid-enhancing factor” activity of extracts from amyloid-containing tissues in AA amyloidosis animal models. Recent data indicate that this activity can reside in macrophages.126


Distribution of Amyloid: Localized Versus Systemic Amyloidosis

Tissue localization of the deposits is characteristic of many amyloidoses (Table 60.7). Single-organ involvement may reflect either local secretion or particular tropism of the amyloid precursor. Systemic amyloidoses are derived from
circulating precursors, which either display unusual structural features or are present at abnormally high plasma levels, or both. Although most cases of LC amyloidosis are due to systemic organ deposition of LCs, localized forms of LC amyloidosis have also been reported mostly in the orbit, larynx, nasopharynx, lung, skin, and the genitourinary tract. A local infiltration of plasma cells is then usually found in proximity to the amyloid deposits, and may be responsible for the secretion of an amyloidogenic LC.


Pathologic Data with Special Emphasis on Renal Involvement

Despite the diversity of amyloidogenic proteins, they all deposit in tissue as fibrils constituted by the stacking of β -pleated sheets as identified by X-ray crystallography and diffraction studies. This unique protein conformation is responsible for the tinctorial and optical properties revealed by Congo red staining of tissue sections, and for the relative resistance of the fibrils to solution in physiologic solvents and to normal proteolytic digestion, which leads to their implacable accumulation in tissues.116

By light microscopy, the deposits are extracellular, eosinophilic, and metachromatic. After Congo red staining, they appear faintly red and show the characteristic applegreen birefringence under polarized light. This light microscopic method is the most reliable to detect amyloid because it yields virtually no false-positive findings. Sections thicker than those usually recommended for renal pathologic examination (i.e., >5µm in thickness) may be necessary to produce sufficient color density. Metachromasia is also observed with crystal violet, which stains the deposits in red. The use of other stains such as thioflavine T has been proposed, but the results lack specificity. The permanganate method may help to discriminate AA from AL fibrils if the sections are treated with permanganate before the Congo red procedure. AL amyloid is resistant, whereas AA amyloid is sensitive to permanganate oxidation. However, this method has been supplanted by immunohistochemical analysis of the deposits.

In the kidney, the earliest lesions are located in the mesangium, along the glomerular basement membrane, and in the blood vessels (Fig. 60.10C). Within the mesangium, deposits are associated primarily with the mesangial matrix, and subsequently irregularly increase by spreading from lobule to lobule and then invading the whole mesangial area. Amyloid deposits may also infiltrate the capillary basement membrane or be localized on both sides of it. When subepithelial deposits predominate, spikes recalling those seen in membranous glomerulopathy may be observed. It was shown that the severity of proteinuria correlated with the presence of spicules and podocyte destruction rather than with the amount of amyloid in the glomerulus. Glomerular cell proliferation is infrequent. Advanced amyloid typically produces a nonproliferative, noninflammatory glomerulopathy, responsible for a marked enlargement of the kidney. The amyloid deposits replace normal glomerular architecture with loss of cellularity. When glomeruli become massively sclerotic, the deposits may be difficult to demonstrate by Congo red staining, and electron microscopy may then be helpful. The latter may also be required at very early stages, which may not be detected by light microscopy examination in patients presenting with the nephrotic syndrome. Except in fibrinogen A α -chain amyloidosis, which characteristically does not affect renal vessels, the media of the blood vessels is prominently involved at early stages. Vascular involvement may predominate, and occasionally occur alone, particularly in AL amyloidosis (Fig. 60.10D). Deposits may also affect the tubules and the interstitium, leading to atrophy and disappearance of the tubular structures and to interstitial fibrosis. In apolipoprotein AI amyloidosis related to the Leu160Pro variant, deposits markedly predominate in the interstitium, whereas glomeruli are not or are occasionally involved.127

Because of the heterogeneity of amyloidotic diseases, which results in specific diagnostic and therapeutic strategies adapted to the type of protein deposited within tissues, immunofluorescence examination of snap-frozen biopsy specimens with specific antisera should be routinely performed.128,129,130 In the first series published by Gallo et al.,129 immunohistochemical classification of amyloid type was possible for 44 (88%) of 50 patients using anti-LC and anti-AA antisera. However, Noel et al.128 pointed out that immunofluorescence with sera directed against HCs and LCs of Ig might be more difficult to interpret than with anti-AA antiserum. This is likely due to the frequent loss of LC constant domains in fibrils, accounting for the absence of epitopes normally recognized by antibodies. It is also possible that the pseudocrystalline structure of the fibrils makes these epitopes poorly accessible to antibodies. In a more recent series, 12 of 34 patients (35.3%) with proven AL amyloidosis had negative immunofluorescence staining for κ and λ light chains, which confirms the relatively low sensitivity of immunofluorescence microscopy in the detection of AL amyloidosis in the kidney and underscores the need to pursue additional diagnostic studies to identify the plasma cell dyscrasia.131

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

May 29, 2016 | Posted by in NEPHROLOGY | Comments Off on Monoclonal Gammopathies: Multiple Myeloma, Amyloidosis, and Related Disorders

Full access? Get Clinical Tree

Get Clinical Tree app for offline access