Membranous Nephropathy

Anne M. Kouri

Sonia Boyer-Sauvet

Michelle M. O’Shaughnessy

INTRODUCTION/GENERAL CONSIDERATIONS

Membranous nephropathy (MN) is a leading cause of nephrotic syndrome in adults without diabetes, and the estimated incidence is 8 to 10 cases per 1 million.¹,² MN is uncommon in children, in whom it accounts for between 1% and 7.5% of children with nephrotic syndrome.³,⁴,⁵,⁶,⁷,⁸ However, it accounts for a larger proportion in steroid-resistant cases and in adolescents and young adults.⁷,⁹,¹⁰,¹¹,¹² MN affects all ethnic and racial groups and both males and females, although Caucasian males older than 40 years of age are disproportionately affected.¹,² The disease is considered “primary” (ie, a kidney-limited, immune-mediated disease) in 70% to 80% of cases, and “secondary” to conditions such as malignancy, drugs, infection, or systemic autoimmune diseases in the remainder.¹³,¹⁴,¹⁵

The pathogenesis of primary MN remained elusive for decades. Deposition of immune complexes in the subepithelial space supported an immune-mediated process, which was reproducible in animals by immunizing susceptible rats with glomerular antigens or antibodies (“Heymann nephritis”).¹⁶,¹⁷ However, the past 20 years have seen an explosion of discovery with the identification of numerous podocyte autoantigens implicated in disease pathogenesis.¹⁸

PATHOGENESIS

Histologic Findings

The term MN describes a histologic pattern of injury by light microscopy, that is, diffuse glomerular basement membrane (GBM) thickening without significant hypercellularity.¹⁹,²⁰ “Spikes” of GBM (ie, new GBM forming around immune deposits) are characteristically observed by silver stain. Glomerular sclerosis and tubulointerstitial fibrosis appear as the disease progresses (Figure 8.1).

Immunofluorescence (IF) microscopy reveals a diffuse granular pattern of immunoglobulin (Ig) G and C3 staining along the GBM, representing immune deposition in the subepithelial space.¹⁹ In the modern era, staining for phospholipase A2 receptor (PLA2R) is also becoming routine practice,²¹,²² whereas staining for IgG subclasses can further help distinguish primary from secondary MN²³,²⁴,²⁵: Predominant IgG4 staining supports primary MN, whereas predominant IgG1, IgG2, or IgG3 staining suggests a secondary cause.

By electron microscopy (EM), subepithelial electron-dense deposits on the outer aspect of the GBM, effacement of overlying podocyte foot processes, and expansion of the GBM by deposition of new extracellular matrix between deposits are observed.¹⁹,²⁰

Histologic Staging in Membranous Nephropathy

Four ultrastructural stages of MN were described by Churg and Ehrenreich almost 50 years ago,²⁶ and this staging system is variably applied in the modern
era: Stage I is characterized by small immune-complex-type electron-dense deposits in the subepithelial space; stage II is characterized by projections of GBM around the subepithelial deposits; stage III is characterized by GBM entirely surrounding the deposits; and stage IV is characterized by loss of electron density of the deposits, resulting in electron-lucent zones within an irregularly thickened GBM. Despite its descriptive utility, this staging system correlates poorly with clinical disease severity or responsiveness to treatment; instead, generic histologic features of disease chronicity, including glomerular scarring and interstitial fibrosis, have been associated with kidney survival.²⁰

FIGURE 8.1: Histopathology of membranous nephropathy. A, Periodic acid-Schiff staining shows marked global thickening of the glomerular basement membrane (GBM), which has a vacuolated appearance (×400). B, Jones methenamine silver staining exhibits prominent GBM spike formation with intervening silver-negative subepithelial deposits (×400). C, Immunofluorescence staining for immunoglobulin G reveals intense, granular global subepithelial positivity involving the glomerular capillary walls (×400). D, Electron microscopy shows global subepithelial electron-dense deposits separated by intervening GBM spikes (×6,000). (Adapted, with permission, from: Markowitz GS, D’Agati VD. Membranous glomerulonephritis. In: Jenette JC, D’Agati VD, Olson JL, Silva FG, eds. Heptinstall’s Pathology of the Kidney. 7th ed. Wolters Kluwer; 2014:255-299.)

Histologic Features Distinguishing Secondary From Primary Membranous Nephropathy

Certain histologic features suggest an underlying secondary cause for MN²⁷,²⁸:

Mesangial, subendothelial, and/or tubular basement membrane immune deposits
A predominance of isotypes other than IgG4, for example, IgG1 and IgG3 predominance in class V lupus nephritis, or IgG1 and IgG2 predominance in malignancy-associated MN
The presence of tubuloreticular inclusions (interferon footprints) in glomerular endothelial cells, supporting class V lupus nephritis

The Role of Specific Podocyte Antigens

Phospholipase A2 Receptor

PLA2R was first described as a major autoantigen in MN in 2009 (Visual Abstract 8.1), being detectable in 70% to 80% of cases of primary MN in adults (Table 8.1).²⁹ PLA2R was originally thought to associate only with primary MN, but anti-PLA2R antibodies have since been detected in cases of MN associated with hepatitis B virus,³⁰ sarcoidosis,³¹ “immunodysregulation, polyendocrinopathy, enteropathy, X-linked (IPEX)” syndrome,³² and malignancy.³³,³⁴ Whether these associations are causative or coincidental remains to be determined. PLA2R is a 180-kDa transmembrane glycoprotein constitutively expressed on the surface of podocytes and composed of a very large extracellular region containing multiple cysteine disulfide bonds (cysteine-rich domain [CysR]), a fibronectin-like type II domain, and eight repeated C-type lectin-like domains (CTLDs).²⁹ PLA2R is also expressed in the lungs. Its high affinity for secreted phospholipase A2 plays a role in lipid mediator synthesis, and recent studies suggest that PLA2R1 also acts as a tumor gene suppressor.³⁵

PLA2R is not expressed by rodent or other animal podocytes, and thus the pathogenic mechanisms underlying PLA2R-associated MN have been slow to elucidate. Passive administration of rabbit anti-PLA2R antibodies to transgenic mice expressing murine PLA2R induces features of the nephrotic syndrome accompanied by small subepithelial deposits³⁶; however, human anti-PLA2R antibodies do not react with mouse PLA2R and cannot be used in this model. At the initiation of the humoral response, podocytes begin to express PLA2R and complement inhibitors at the cell surface; circulating autoantibodies then target PLA2R on the basal aspect of podocytes. Ongoing synthesis and delivery of PLA2R to the cell surface enable continued immune deposit formation in the presence of circulating autoantibodies.³⁷,³⁸ Supporting this pathogenic model, detection of PLA2R antigens in kidney biopsy specimens using IF or immunohistochemistry appears to be a specific diagnostic tool for MN,²¹,²⁷ whereas detection of circulating anti-PLA2R antibodies can allow for the noninvasive diagnosis of MN and for monitoring of disease activity and treatment response.³⁹,⁴⁰,⁴¹,⁴² Recent studies have also demonstrated the occurrence of epitope spreading in cases of PLA2R-associated MN. Epitope spreading describes the diversification of epitope specificity from the initial epitope-specific immune response to additional subdominant and/or cryptic epitopes on that protein. Primary epitopes for PLA2R have been described in the CysR domain and in the CTLD1, CTLD7, and CTLD8 domains.⁴³,⁴⁴ PLA2R epitope spreading at baseline appears to predict a reduced likelihood of remission of MN and a poorer prognosis.⁴³,⁴⁵

Thrombospondin Type-1 Domain-Containing 7A

Thrombospondin type-1 domain-containing 7A (THSD7A) is a transmembrane protein expressed on podocytes and implicated in 3% to 5% of cases with MN.⁴⁶ It may be more prevalent in Japanese compared to Caucasian patients with primary MN.⁴⁷ THSD7A has additionally been identified in cases of malignancy-associated MN,⁴⁸ with tumor cells also expressing THSD7A. No immunodominant epitope has been described.⁴⁹ Circulating anti-THSD7A antibody levels appear to correlate with disease activity and might become a useful clinical biomarker.⁵⁰

Anti-THSD7A antibodies passively transferred from humans with MN into mice can induce the clinical and histologic features of MN.⁵⁰,⁵¹

TABLE 8.1 Antigens Associated With Membranous Nephropathy

	Primary vs Secondary MN Disease Associations	Circulating Antibody Identified	Immune Deposits Typically Identified by Immunofluorescence Microscopy
PLA2R	Primary MN, rarely secondary MN	Yes	IgG4, C3
THSD7A	Primary MN, malignancy in up to 20% of patients	Yes	IgG4, C3
EXT1 and EXT2	Autoimmune disease (SLE, MCTD) in most patients	No	IgG1, IgA, IgM, C3, C1q
NELL-1	Primary MN, malignancy in up to 30% of patients	Yes	IgG1, C3
PCDH7	Primary MN but sometimes associated with autoimmune diseases and malignancy	Yes	IgG1-4, C3, C1q
NCAM-1	SLE in most patients	Yes	IgG1 ± other subclasses, IgA, IgM, C1q, C3
Sema3B	Children, positive family history	Yes	IgG1, C3 + TBM deposits
HTRA1	ANCA, malignancy	Yes	IgG4
ANCA, antineutrophil cytoplasmic autoantibodies; EXT1 and EXT2, exostosins 1 and 2; HTRA1, high-temperature recombinant protein A1: IgG, immunoglobulin G; MCTD, mixed connective tissue disease; MN, membranous nephropathy; NCAM-1, neural cell adhesion molecule 1; NELL-1, neural epidermal growth factor-like 1; PCDH7, protocadherin 7; PLA2R, phospholipase A2 receptor; Sema3B, Semaphorin 3B; SLE, systemic lupus erythematosus; TBM, tubular basement membrane; THSD7A, thrombospondin type-1 domain-containing 7A.

Neural Epidermal Growth Factor-Like 1

Neural epidermal growth factor-like 1 was recently identified in a substantial minority of cases of anti-PLA2R-negative MN, both in presumed primary MN and in malignancy-associated MN.¹⁸,²⁶

Exostosin 1/Exostosin 2

Exostosin (EXT) 1 and EXT2 deposits were also recently detected in patients with MN secondary to autoimmune diseases, particularly systemic lupus erythematosus.⁵²

Semaphorin 3B (Sema3B)

Semaphorin 3B (Sema3B) is predominantly identified in children (younger than 2 years) and in young adults with MN.⁵³

Others

Protocadherin 7,⁵⁴ High-Temperature Recombinant Protein A1,⁵⁵ and neural cell adhesion molecule 1⁵⁶ are additional putative autoantigens in MN.

The Role of T Cells

Many studies have reported a dysregulated immune phenotype in MN characterized by a reduction in regulatory T cells (Treg) and polarization toward a Th2-dominant immune response.⁵⁷,⁵⁸,⁵⁹ Recent studies also implicate Th17 activation and an imbalance between Th17 and Treg cells in disease pathogenesis,⁶⁰,⁶¹,⁶² whereas others have demonstrated that an early increase in Treg cells after treatment with rituximab (RTX) predicts an earlier treatment response.⁶⁰

The Role of B Cells

A key role for B cells in MN, both as autoantibody-producing cells and as antigen-presenting cells,⁶³,⁶⁴,⁶⁵ provides the rationale for B-cell-targeted therapies.⁶³,⁶⁴,⁶⁵,⁶⁶ Diverse B-cell lineages have been implicated in MN pathogenesis: CD20⁺ activated B cells in the spleen and lymph nodes; CD19⁺ and CD20^– plasmablasts and short-lived plasma cells in the blood; and CD19^–, CD20^–, and CD38⁺ long-lived memory plasma cells in the bone marrow and ectopically in inflamed kidneys.⁶⁷ The latter produce considerable amounts of IgG autoantibodies, yet they lack CD19 and CD20 surface expression, making them largely resistant to anti-CD20 monoclonal antibody therapies.⁶⁸,⁶⁹ This might explain the nonuniversal achievement of sustained complete remission following RTX therapy in patients with primary MN, despite successful depletion of circulating B cells.⁴¹,⁷⁰

The Role of the Complement System

Complement activation resulting in C5b-9 deposition in the GBM has been implicated in the podocyte injury and resulting proteinuria observed in MN. This hypothesis is supported by the experimental Heymann nephritis rat model¹⁶,¹⁷: Without complement-activating IgG, or complement components, subepithelial deposits fail to induce podocyte injury.¹⁷

Evidence for complement activation in human MN is supported by the presence of complement components within subepithelial deposits: C3 staining is typically positive, whereas C1q staining is usually weak or negative.¹⁹ The role of complement pathways upstream of C5 in human and experimental MN remains poorly understood.⁷¹ MN associated with PLA2R or THSD7A antibodies
is characterized by predominant IgG4 with minimal IgG1or IgG3 deposition.⁷²,⁷³ However, IgG4 does not bind C1q and is unable to activate the classical complement pathway.²⁵,⁷⁴,⁷⁵ Instead, the predominance of noncomplement-fixing IgG4 deposition, and paucity of C1q deposition, suggests that the alternative or lectin complement pathways play a bigger role in disease pathogenesis.⁷⁶,⁷⁷ In one cohort, glomerular deposits of C4d, C3d, and C5b-9 were detected in all patients with MN, whereas mannose-binding lectin (MBL) was detected in 40% to 45% of patients,⁷⁶ implicating the MBL complement pathway in MN pathogenesis. However, cases of PLA2R-associated MN in patients with complete MBL deficiency, and predominant alternative complement pathway activation, are reported.⁷⁸ On balance, the relative importance of the classic, alternative, and MBL complement pathways might vary across patients, and all three represent potential therapeutic targets warranting further investigation.⁷¹,⁷⁴,⁷⁹

The Role of Environmental Factors

Environmental factors including air pollution, vitamin D deficiency, and smoking have been reported to act as danger signals, redirecting the immune response toward a Th2 or Th17 phenotype.⁶¹,⁸⁰,⁸¹ Two Chinese studies have demonstrated a strong association between air pollution and MN incidence.⁸²,⁸³ Similar findings have been reported in a French cohort, suggesting that an environmental factor might trigger disease in patients with a Th17 profile.⁶¹

The Role of Genetic Factors

An important Genome-Wide Association Study (GWAS) identified single-nucleotide polymorphisms in two loci that were highly associated with primary MN⁸⁴: the PLA2R locus on chromosome 2q24 and the human leukocyte antigen (HLA) complex class II on chromosome 6, especially HLA-DQA1 and HLA-DR3.⁸⁵,⁸⁶ HLA-D alleles play a role in presenting PLA2R antigen to the immune system and might also impact MN prognosis.⁸⁷ These findings were confirmed in a second large GWAS of 3,782 cases of primary MN and 9,038 controls of East Asian and European ancestry.⁸⁸ This study also identified two new loci highly associated with MN: one in NFKB1 and another in IRF4, both encoding molecules involved in immune regulation.

CLINICAL FINDINGS

Signs and Symptoms

Most patients (80%) present with nephrotic syndrome, defined as proteinuria greater than 3.5 g/d, a serum albumin less than 35 g/L by bromocresol green (or <30 g/L by bromocresol purple or immunonephelometry), and the presence of edema. At presentation, more than 70% of patients have normal blood pressure.⁸⁹ Rare cases of crescentic MN, presenting with proteinuria, hematuria, and rapid decline in kidney function (in the absence of antineutrophil cytoplasmic antibodies or anti-GBM antibodies) have been reported.⁹⁰ MN can also first present with a thromboembolic complication (8% of patients).⁹¹

Laboratory Findings

Proteinuria ranges from subnephrotic to more than 20 g/d.⁸⁹ Marked proteinuria without hypoalbuminemia is rare. Microscopic hematuria occurs in up to 50%⁹² of cases but macroscopic hematuria is rare. Glomerular filtration rate is normal at presentation in 70% of patients, and acute kidney injury (AKI) is uncommon.
When it occurs, it is usually secondary to functional hypovolemia or prescribed drugs (diuretics, renin-angiotensin system blockers), resulting in prerenal azotemia or acute tubular injury. Acute renal vein thrombosis is a differential diagnosis. Severe hyperlipidemia is frequently identified in patients presenting with nephrotic syndrome.⁹³

Imaging Studies and Screening for Secondary Causes

Kidney ultrasound should be performed if kidney function is impaired or kidney biopsy is proposed. In cases of AKI, imaging can also detect a renal vein thrombosis. Screening for secondary causes of MN should be performed in most cases, irrespective of the presence or absence of anti-PLA2R, anti-THSD7A, or other antibodies.⁹⁴ This workup should include (Table 8.2) antinuclear antibodies; serum C3 and C4 complement levels; screening for hepatitis B, hepatitis C, and HIV infection; and (in patients older than 50 years) serum and urine protein electrophoresis with immunofixation and serum immunoglobulin free light chain tests. Patients should also be screened for malignancy according to population and age-appropriate guidelines. The index of suspicion for malignancy or secondary causes of MN should be higher in anti-PLA2R-negative cases and in the presence of atypical histologic features by kidney biopsy (see earlier text).

Special Tests

Circulating anti-PLA2R antibodies can be detected using various assays. The most popular is an enzyme-linked immunosorbent assay (ELISA), which is reported to have 99.6% specificity.⁹⁵,⁹⁶ Indirect IF assays may be more sensitive than is ELISA, but they report an antibody titer rather than a quantitative concentration and are less useful as a disease activity biomarker.⁹⁵,⁹⁶ Analysis of PLA2R epitope spreading is not yet clinically available.⁴³,⁴⁴ Assays for circulating anti-THSD7A are becoming commercially available,⁹⁷ whereas assays for antibodies against other newer podocyte antigens are in development.

Kidney biopsy was traditionally required to confirm a diagnosis of MN, although a positive anti-PLA2R antibody test in the presence of nephrotic syndrome, normal kidney function, the absence of prior immunosuppressive therapy, and low suspicion of a secondary disease driver is considered highly specific for a histologic diagnosis of MN and may obviate the need for an invasive kidney biopsy in adult patients.²²

DIFFERENTIAL DIAGNOSIS

The differential diagnosis of MN includes any glomerular disorder characterized by nephrotic or subnephrotic range proteinuria and/or microscopic hematuria, including minimal change disease (MCD), focal segmental glomerulosclerosis (FSGS), IgA nephropathy, Alport syndrome, renal amyloid, or diabetic nephropathy.

COMPLICATIONS

Complications of MN include the following⁹⁸:

Generic complications associated with chronic kidney disease or AKI, for example, hypertension, electrolyte derangements, anemia
Complications shared with other causes of nephrotic syndrome, for example, hyperlipidemia, increased risk of infection, thromboembolic events, and cardiovascular events

TABLE 8.2 Potential Underlying Causes for Secondary Membranous Nephropathy

Malignancy	Lung Breast Gastrointestinal Prostate (nonexhaustive)
Infection	Viral infections HBV HCV (rare) HIV (rare) Bacterial infections Syphilis Streptococcal infections (rare) Parasitic infections Schistosomiasis Malaria Leprosy Filariasis
Immune diseases	SLE (class V lupus nephritis) Sjögren syndrome Sarcoidosis IgG4-related disease Rheumatoid arthritis (rare) Autoimmune thyroiditis Urticarial vasculitis
Drugs	NSAIDs Penicillamine Gold salts Mercury Captopril Clopidogrel Solvents Anti-TNF therapies
Hematopoietic stem cell transplant/graft vs host disease
HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; IgG, immunoglobulin G; NSAIDs, nonsteroidal anti-inflammatory drugs; SLE, systemic lupus erythematosus; TNF, tumor necrosis factor.