Professors Ronco and Debiec from the Sorbonne recently published a review of new research and treatment advances in membranous nephropathy in The Lancet, summarizing new developments in the pathogenesis, diagnosis and treatment of membranous nephropathy (MN) over the past 13 years. This article compiles the more clinically relevant parts of the review for our colleagues to study together.
Epidemiological and clinical features of membranous nephropathy
The annual incidence of MN in adults is 1 new case/100,000/year worldwide, or 10,000 new cases per year in Europe.MN affects patients of all ages and ethnicities, but is more frequent in men than in women (sex ratio = 2:1), with a peak incidence in the 30s and 50s. The possibility of MN in young women is highly suspect for lupus. MN is uncommon in children (<5% of renal biopsy specimens from children) and is often associated with hepatitis B and less frequently with autoimmune or thyroid disease.
The remainder present with asymptomatic proteinuria (<3.5 g/24 h), of which 60% progress to full-blown nephrotic syndrome. Microscopic hematuria occurs in 50% of MN patients, but massive hematuria and erythrocyte tubularity are rare. 80% of MN patients have normal blood pressure and glomerular filtration rate at presentation. Acute kidney injury is rare and may result from hypovolemia due to excessive diuresis, acute bilateral deep vein thrombosis, drug-induced interstitial nephritis or superimposed crescentic glomerulonephritis.
MN is a chronic disease that can resolve spontaneously and recur. It often resolves spontaneously in approximately 40% of cases within the first 2 years of disease onset. Predictors of spontaneous remission are proteinuria levels <8 g/d at baseline, female age <50 years, and good renal function at the time of onset. MN remains the second or third leading cause of end-stage renal disease (ESRD) in primary glomerulonephritis.
Pathological changes of membranous nephropathy
In the early stages of MN, the glomerular structure is normal on light microscopy and the diagnosis can only be made by immunofluorescence and electron microscopy. The next stage is characterized by uniform thickening of the capillary wall. Early reflections of deposits between the glomerular basement membrane are called spikes. As the disease progresses, the deposits merge into the glomerular basement membrane and appear as if the immune deposits have been absorbed. As the disease progresses further, interstitial fibrosis and glomerulosclerosis emerge.
The diagnosis of both idiopathic and secondary MN relies on the finding of IgG granule-like deposits in the glomerular capillary collaterals. The presence of C1q suggests secondary MN, especially in association with SLE. In addition, biopsy specimens should be routinely stained for PLA2R antigen and IgG subtypes. IgG4 is predominant in idiopathic MN, while IgG1 and IgG2 subtypes are predominant in secondary MN.
Electron microscopy shows significant injury with deposition of subepithelial electron dense material and fusion of podocyte peduncles. Disease can be staged according to the extent to which subepithelial immune complexes are surrounded by the glomerular basement membrane.
Identification of target antigens and risk genes for human membranous nephropathy
Major advances in the study of MN pathophysiology began in the early 2000’s. In 2002, podocyte neutral endopeptidase was identified as a target antigen for circulating antibodies in alloimmune neonatal nephropathy. Neutral endopeptidase-associated alloimmune glomerulopathy defined a novel organ-specific disease caused by maternal-infant incompatibility, and although this disease was rare, analysis of its pathogenesis provided evidence for the notion that podocyte antigens could cause MN in humans, laying the groundwork for later identification of PLA2R involvement in adult MN.
The search for idiopathic MN target antigens was unsuccessful for many years thereafter. Between 2009 -2014, two podocyte proteins were identified through the use of human glomerular microdissection, metabolomics techniques and high throughput technologies.
The first major autoantigen was PLA2R type M. Autoantibodies against PLA2R were detected in the circulation of 70% of patients with idiopathic MN. The second autoantigen was THSD7A, and autoantibodies against this protein were detected in the circulation of 5-10% of idiopathic MN patients negative for anti-PLA2R. PLA2R and THSD7A were detected in normal human glomeruli, and both antigens were co-localized with IgG4 within the subepithelial deposits.
In addition, IgG eluting from biopsy specimens reacts with recombinant PLA2R and THSD7A, which have similar structural and biochemical properties. The autoantibodies that recognize these two antigens are mainly the IgG4 subtype. Interestingly, patients with idiopathic MN have either an autoimmune response to PLA2R or to THSD7A, but not to both antigens. This suggests that PLA2R-associated MN and THSD7A-associated MN are two separate molecular mechanisms and that these antigens are the primary targets of idiopathic MN.
The non-pedunculated antigens are the main effectors of secondary MN. Cationized bovine serum albumin (BSA) was identified as the primary antigen in children younger than 5 years of age, with high titers of serum anti-BSA antibodies (IgG1 and IgG4). If BSA is a common dietary antigen due to early exposure to bovine dairy products, then other dietary and environmental antigens may also contribute to disease by similar mechanisms. Enzymes used in replacement therapy for enzyme deficiency in the treatment of lysosomal storage disease, such as hepatitis B and C viruses, may also be involved in secondary MN.
Although MN is not a typically inherited disease, the influence of genetic factors has been confirmed in disease models in rats and mice, as well as in European Caucasians, showing a strong association between haplotype HLA-B8DR3 and other HLA class II immune response genes and MN. Genome-wide association (GWAS) studies have reported a high correlation between the 6p21 HLA-DQA1 and 2q24PLA2R1 loci and idiopathic MN in European Caucasians.
Diagnosis of Membranous Nephropathy
Over the last 3 years, many studies on the incidence of anti-PLA2R antibodies in different pathological types have shown these antibodies to be biomarkers of MN specificity and sensitivity. In a 2014 meta-analysis that included 9 articles and included 15 studies with a total of 2212 patients, anti-PLA2R antibodies had a specificity of 99% (95% CI 96-100) and a sensitivity of 78% (95% CI 66-87). Anti-PLA2R antibodies were not found in other renal or autoimmune diseases or in healthy individuals.
The specificity of anti-PLA2R antibodies is so high that some clinicians are reconsidering the utility of renal biopsy, especially in older patients who are less well and in patients with life-threatening complications (such as renal vein thrombosis or pulmonary embolism) requiring anticoagulation.
The incidence of anti-PLA2R antibodies is low in secondary MN, but it is difficult to exclude a coincidence of MN with associated disease. In secondary MN, the incidence of anti-PLA2R antibodies associated with hepatitis B virus replication and active nodal disease is elevated, suggesting that these two diseases associated with immune disorders may induce or enhance the immune response against PLA2R.
PLA2R antigen can also be detected in the sediment in the absence of circulating antibodies. There are several explanations for this, including rapid clearance of antibodies from the blood due to high affinity for PLA2R, immune remission, or renal biopsy performed long after the onset of the disease. Detection of antigens in immune complexes has made it possible to retrospectively study PLA2R-associated MN. Conversely, in some patients, circulating anti-PLA2R antibodies do not correlate with PLA2R in the sediment, suggesting that these antibodies may not be pathogenic.
Combined serum (antibody) and biopsy tissue (antigen) analysis is recommended in all patients with MN. Detection of PLA2R antigens in immune complexes is also an important clue for the diagnosis of primary MN, where PLA2R is usually associated with IgG4-based deposits. However, the presence of PLA2R in immune complexes is usually reported in patients with replicating hepatitis B virus infection and active nodal disease.
Predictive value of anti-PLA2R antibodies
Many studies performed in the last 3 years have shown that anti-PLA2R antibody levels correlate with urinary protein excretion and disease activity. Antibody levels are undetectable in spontaneous remission or treatment-induced remission and reappear or increase when the disease relapses. Antibody levels also predict prognosis, as high titer antibody levels are associated with a decreased risk of spontaneous remission or immunosuppression-induced remission and with an increased risk of developing nephrotic syndrome and worsening renal function in patients without nephrotic syndrome. The time interval between initiation of immunosuppressive therapy and remission was significantly increased in patients in the highest antibody titer group.
Patient follow-up and monitoring of treatment effectiveness
Several studies have shown partial or complete reduction of anti-PLA2R antibodies weeks or months before remission of renal disease. In a collaborative study that included 132 patients given rituximab for severe nephrotic syndrome, 81 patients were positive for anti-PLA2R antibodies, and five important findings were identified.
First, while remission rates in patients with positive anti-PLA2R antibodies were negatively associated with antibody titers, remission rates were similar in patients with PLA2R-associated or unassociated MN. Second, in a multifactorial analysis, complete remission of antibodies at 6 months was strongly and independently associated with remission of renal function. Third, complete remission tended to occur prior to complete antibody depletion. Fourth, antibody titer reduction occurred approximately 10 months prior to the reduction in urinary protein excretion. Fifth, re-emergence or elevation of antibodies occurs prior to renal disease recurrence. The interval of several months from immune remission to renal remission may be explained by deposit remodeling and glomerular capillary wall recovery.
In addition, antibody titers at the end of immunosuppression predict the onset of disease relapse. Five years after immunosuppression discontinuation, 58% of patients with anti-PLA2R antibody clearance did not relapse, whereas relapse tended to occur in those patients who still had antibodies at the end of treatment. In conclusion, the presence or absence of anti-PLA2R antibodies does not predict response to immunosuppressive therapy, but immune remission is a strong predictor of clinical remission in patients with PLA2R-associated MN.
Three important questions are listed and should be further confirmed by randomized controlled trials. First, should we treat patients with higher and lower titers earlier than those with lower titers? Second, should we start immunosuppressive therapy when antibody titers are still high after 3-6 months? Finally, should we continue treatment when antibody titers re-emerge or increase?
We recommend testing antibody titers every 2 months prior to starting immunosuppressive therapy to avoid giving patients unnecessary treatment when they enter immune remission. Antibody titers should be tested every month for the first 6 months of immunosuppressive therapy.
Monitoring of renal transplant patients
Due to the high recurrence rate of MN after renal transplantation, frequent antibody testing should be initiated from the date of transplantation. 50-80% of recurrent MN and all early relapses are associated with anti-PLA2R antibodies. These findings and atypical MN recurrences with monoclonal IgG3κ deposition support the pathogenic role of anti-PLA2R antibodies.
In contrast, certain patients with high anti-PLA2R antibody titers never relapsed, even histologically. The risk of recurrence appears to be increased when the patient’s IgG4-type anti-PLA2R antibody titer does not decrease after transplantation. In ab initio MN, which may be an alloimmune disease, detection of anti-PLA2R antibodies in the serum and PLA2R antigen in the sediment is almost always negative.
From pathophysiological progression to new therapeutic pathways
The pathological mechanisms of MN over the past 13 years have opened new doors for treatment. First, anti-PLA2R antibody titers have an important role in selecting patients for immunosuppressive therapy, the first step in individualized treatment. Second, non-specific immunosuppressive therapy is replaced or combined with more specific and less toxic therapies. Although 1/3 of patients with persistent nephrotic syndrome do not respond to therapy, anti-CD20 antibodies (rituximab) are still a step in this right direction.
More specific epitope-driven therapies based on specific immunoadsorption are also needed. Growing knowledge of the molecular structure of so-called nephritic antigenic epitopes has enabled researchers to design non-peptide receptor antagonists to be used as decoys for antibody decay. Third, clinicians should always keep in mind that even if the immune response is rapid in responding patients, full activation and damage to podocytes persists until antibody reduction is complete.
Therefore, there are still opportunities to develop complement antagonists that block the formation of C5b-9 membrane attack complexes and cytoprotective therapies that facilitate glomerular capillary wall recovery. Research in this area should be intensified.
Conclusion
Over the past 13 years, advances in the study of the molecular pathogenesis of membranous nephropathy have provided clinicians with sensitive and specific biomarkers (e.g., anti-PLA2R antibodies in serum and PLA2R antigens in immune deposits), opening the door to individualized drug therapy. These biomarkers will be necessary for nephrologists when designing future clinical trials for membranous nephropathy.