Selection Committee Member: Shree Sharma @nephrosharma
Shree Sharma is the Laboratory Director at Arkana Laboratories in Little Rock, AR. He has published multiple articles and book chapters and is passionate about educating residents and fellows. Dr. Sharma has won numerous awards for his outstanding teaching, including the Top 5 of the 40 Under 40 awardees organized by the American Society of Clinical Pathology. His passion is to advance the field of kidney pathology.
Writer: Elena Cervantes @Elena_Cervants
Elena Cervantes is an academic nephrologist at Johns Hopkins University in Baltimore, MD, where she completed her fellowship as part of the Clinical Educator Pathway. She is the co-director of the medical student Nephrology course in the School of Medicine and a graduate of the 2021 class of the Nephrology Social Media Collective (NSMC) Internship. Her focus is on developing meaningful, creative educational curricula.
Competitors for the Nephropath Region
PLA2R vs Other Membranous Antigens
One of the most common causes of nephrotic syndrome among adults is membranous nephropathy (MN). Traditionally, the disease has been classified into “primary”/“idiopathic” and “secondary” to systemic conditions or drugs. The Heymann nephritis rat model in 1959 established the autoimmune nature of the disease. Approximately two decades later, Couser et al in the United States and Van Damme et al in the Netherlands demonstrated that the immune complex deposits were formed in situ along the glomerular basement membrane (GBM). Later in the 1980s, the rat antigen was identified as megalin. The human antigen remained a mystery until 2009, when Beck, Salant, and colleagues recognized the M-type phospholipase A2 receptor (PLA2R) as the antigenic target in humans. This groundbreaking antigen was present in 70-80% of the cases of primary MN and opened doors in the field as a prognostic and diagnostic marker, and hopefully even a potential target for future therapy.
More recently, laser microdissection and mass spectrometry led to the identification of other multiple target antigens such as THSD7A, EXT-1/EXT2, NCAM-1, NELL1, Sema3B, HTRA1, PCDH7, TGFBR3, and CNTN1, with even more in the pipeline. In fact, these proteins soon gained clinical relevance as different conditions appeared to be specific to each of them.
This matchup pits a bona fide superstar in PLA2R against all of the up-and-coming freshmen that are ready to make the jump to the pros!
Just by virtue of being the first human antigen that created a breakthrough in our understanding of primary MN, this antigen might be a personal favorite for several readers. It is the de facto king of the court, and now has over a decade of experience under its belt.
PLA2R is a transmembrane protein of 185 kDa of the mannose receptor family. It is found in the external surface of healthy human podocytes, but its original function remains unclear. It has been suggested that the immune system recognizes PLA2R as an autoantigen only when the epitope is exposed in a conformation-dependent manner. Furthermore, epitope spreading has been associated with active disease and poor renal prognosis. On a brighter side, the knowledge we have gained through the identification of these major epitopes could enable antigen specific therapies for this condition.
The kidney biopsy tissue under immunofluorescence (IF) of PLA2R-positive MN co-localizes with immunoglobulin G (IgG), C3, and light chains in a fine granular pattern along the GBM. Additionally, serum antibodies against PLA2R can be measured by enzyme-linked immunosorbent assay (ELISA) and/or by indirect IF, which has enabled a broader understanding of the disease along with real world clinical applications. These antibodies have been found in approximately 70-80% of patients with primary MN and almost 60% of all MN. In fact, a serology-based approach to MN has been proposed. This model illustrates the “kidney-as-a-sink” hypothesis, which explains the low detection of circulating antibodies in initial stages due to the high affinity of the antibodies to the in situ antigens.
The discovery of the PLA2R antigen from the animal model of Heymann nephritis has been recognized as a great example of translational research. For example, kidney biopsy continues to be the gold standard for the diagnosis of MN. However, resources and complications (particularly in high risk populations) have led us to reconsider this practice. Meta-analyses of large cohorts have shown a high sensitivity (78%) and specificity (99%) of the serum PLA2R antibody in this disease. These findings have moved our field forward, and a non-invasive diagnosis of MN has been proposed based on ELISA and immunofluorescence assay (IFA) measurements. In this study, ELISA values >20 RU/mL alone, or ELISA values ≥ 2 RU/mL and <20 RU/mL with a positive IFA confirm the diagnosis of MN without the need for a kidney biopsy Importantly, this criteria can only be applied to patients with preserved kidney function (eGFR > 60 mL/min/1.73 m2) with negative secondary workup for MN and no diabetes.
This non-invasive diagnosis is also supported by the KDIGO 2021 Guidelines for the Management of Glomerular Diseases, which use slightly different ELISA cut-offs for a noninvasive diagnosis of disease (ELISA > 14 RU/mL, or ELISA between 2 and 14 RU/mL with a positive IFA). We should keep in mind that these cut-offs have not been validated and that a biopsy may still be needed if there is a poor response to initial immunosuppressive therapy.
How about using PLA2R for prognosis and deciding whom to treat? All patients should receive supportive therapy with angiotensin-converting enzyme inhibitors (ACE-i) or angiotensin receptor blockers (ARBs), and anti-PLA2R titers should be followed at 3-6 month intervals to estimate risk of progression and response to therapy. Traditionally, we have relied on proteinuria as our primary prognostic marker, with highly nephrotic patients universally needing treatment, and sub-nephrotic patients being considered for supportive therapy alone. In this cohort study, anti-PLA2R levels between 150–300 RU/mL had a positive predictive value of 77% for progression of kidney disease, which suggest that patients with high levels have a low risk of spontaneous remission and warrant up-front immunosuppressive therapy. The KDIGO 2021 Guidelines now include PLA2R antibody titers as part of the criteria to establish the different risk categories and guide therapy (levels > 50 RU/mL and > 150 RU/mL define high risk and very high risk MN, respectively).
Moving on to therapy, studies have looked at using PLA2R for response to therapy and monitoring for remission. Ruggenenti et al showed that complete PLA2R antibody depletion was observed in 72% of patients with MN treated with rituximab. Of these patients, 89% had at least partial remission compared to 11% of patients who still had detectable PLA2R levels. Importantly, depletion of PLA2R levels preceded clinical remission (decrease in proteinuria and increasing serum albumin) by 10 months, again reinforcing its utility as a prognostic marker. Because the duration of B-cell depletion is variable with rituximab, patients who had re-emergence of PLA2R antibodies in their serum were 10x more likely to relapse, and the rise in serum PLA2R preceded clinical relapse by 2.6 months.
Lastly, how about the use of this marker in transplant? Even if patients progress to end-stage kidney disease (ESKD), the titers matter! Anti-PLA2R levels >45 RU/mL pre-transplant have been associated with increased risk of recurrence after transplantation (cutoffs haven’t been validated). The diagram below summarizes risk stratification and therapy approaches based on PLA2R antibody levels.
Other Membranous Antigens
What about the “newer” antigens?
Just because PLA2R has been around since 2009 doesn’t mean that it is a shoo-in champion. There is strength in numbers and we’re seeing quite a few competitor antigens that are teaming up against the reigning champ.
The first is Thrombospondin type 1 domain-containing 7A (THSD7A), a 250 kDa transmembrane protein expressed just beneath the slit diaphragm of healthy glomerular podocytes. It has been found in 3%-5% of the cases of PLA2R-negative primary MN. Similar to its predecessor, autoantibodies recognize one or more conformation-dependent epitopes and antibody levels also correlate with disease activity. In terms of IF in kidney biopsy samples, this antigen shows a granular glomerular capillary wall expression along with IgG, C3, and light chains. From a clinical perspective, THSD7A antigen had slight male predominance, but no further differences with PLA2R were evident at that time.
In 2016, two years after its discovery, Hoxha et al located THSD7A in gallbladder and lymph node tissues of a patient with metastatic primary gallbladder carcinoma and secondary MN. The authors further identified a malignant tumor in 7 out of 25 patients with MN and THSD7A antibodies. Additionally, a retrospective multicenter cohort study determined different clinicopathologic phenotypes based on podocyte antigen staining and showed that THSD7A positivity carried a higher cancer risk. Perhaps if one reflects on the importance of this discovery, one might realize that the clear line between “primary” and “secondary” MN starts to blur. Therefore, in the future, antigen-based classification might be more practical for clinical use.
Moving on to 2019, Sethi et al identified Exostosin-1/Exostosin-2 (EXT-1/EXT2), an antigen that was expressed in young females with MN and positive autoimmune serologies. Almost 85% of the patients had features suggestive of autoimmune conditions on kidney biopsy, with lupus being the most common one, followed by mixed connective tissue disorders. Furthermore, IF revealed a granular staining along the GBM, similar to the staining pattern of PLA2R. An important difference however, was the co-localization with IgG1. Another point of note was that lupus nephritis without a membranous component, did not exhibit EXT1/EXT2 positivity. Interestingly, EXT are inner proteins located in the endoplasmic reticulum and it is thought that EXT1/EXT2 may be involved in the synthesis of one of the components of the GBM, heparan sulfate. To date, antibodies against EXT1/EXT2 haven’t been identified.
Common among older patients and without sex predilection, neural EGF-like-1 protein (NELL-1) was discovered among 16% of PLA2R negative biopsies. This made it the second most common antigen after PLA2R. Additionally, an absence of secondary features such as infections or autoimmune conditions was observed. However, a subset of patients developed a malignancy months after the MN diagnosis, once again challenging the older nomenclature. Similar to EXT1/EXT2, NELL-1 showed granular staining along the GBM and co-localization with IgG1. Gladly, anti-NELL-1 antibodies were identified in response to a conformation-dependent epitope.
Semaphorin 3 is a transmembrane protein that has been located in endothelial cells, podocytes, and tubular epithelial cells. It has been suggested that Sema 3A regulates slit diaphragm proteins and podocyte apoptosis. However, the function of Sema 3B is still unclear. Cohorts in the United States, France, and Italy identified Sema 3B in cases of MN in children <2 years old, and some adult cases of MN with a mean age of 36 years. The authors suggest that pediatric patients who were diagnosed with steroid resistant nephrotic syndrome may actually have MN secondary to antibodies to Sema 3B. Sema 3B is stained predominantly for IgG1 in a granular pattern along the GBM. Tubular basement membrane deposits have been located in some cases, but these deposits were negative for Sema 3B.
The most recent entry into the field is neural cell adhesion molecule 1 (NCAM-1), which was observed in younger females. NCAM1 is predominantly expressed in the nervous system, thyroid and adrenal glands, and immune cells, among others. In the kidneys, NCAM-1 is mainly seen during development and in adults, it has been located in interstitial cells and within podocytes. In MN, NCAM-1 stained positive along the glomerular capillary loops and it co-localized with IgG. IgG subclasses did not show a consistent pattern of staining, although very few cases were IgG4 dominant or co-dominant. Like Sema 3B, tubular basement deposits have been observed, but the majority of them stain negative for NCAM-1. Furthermore, similar to EXT1/EXT2, an association with lupus has been observed. In this report by Caza et al, 40% of cases with lupus nephritis had concomitant neuropsychiatric manifestations such as seizures, cerebritis, serous central retinopathy, psychosis, and stroke.
The list and complexity of these antigens keeps growing. Other rookie proteins that came to play in 2021 include protocadherin 7 (PCDH7). whose distinct feature was a reduced complement activation, and contactin 1 (CNTN1) associated with inflammatory neuropathies. We expect additional players to enter the game in the near future.
The discovery of PLA2R as the target antigen in 70%-80% was just the start of a more thorough investigation to identify the remaining proteins. Despite the progress made in the past decade with PLA2R, we are no closer to a targeted treatment for this challenging disease. This fascinating journey opened a Pandora’s box of “newer antigens” with various associations, unique pathophysiologies, and unknown potential. With each new discovery, we move away from the world of “primary membranous” and one step closer to redefining the disease based on the specific antigen.
DNAJB9 vs IgG4
Traditionally, organized glomerular deposition diseases have been classified into amyloid and non-amyloid, based on Congo red staining. This distinction was thought to determine the cause of the condition, its treatment, and prognosis. The non-amyloid, Congo red negative family includes diseases like fibrillary glomerulonephritis (FGN), immunotactoid glomerulopathy (ITG), and collagenofibrotic glomerulonephritis, whose substructures have classically been defined by the arbitrary and somewhat subjective criteria of fibril diameter.
FGN was first described in 1977 based on distinctive pathologic findings in a patient with nephrotic syndrome. Generally, light microscopy (LM) shows mesangial expansion/hypercellularity with or without duplication of the GBM. Its IF is typically positive for IgG, C3, and kappa and lambda light chains in a smudgy pattern, suggesting immune complex deposition. Electron microscopy (EM) is characterized by the glomerular deposition of randomly oriented fibrils with a diameter of approximately 20 nm within the mesangium and/or the GBM. Some of these characteristics (eg, mesangial expansion on LM or the presence of fibrils on EM) overlap with far more common conditions like amyloidosis and diabetic fibrillosis. What’s that you say, at least we have the Congo red and DNAJB9 stain to help us?
Not so fast! This uncommon condition became even more challenging to diagnose when atypical cases of Congo red positivity were described in 18 patients with FGN. In addition, other rare presentations with negative immunoglobulins by IF have been reported. At this point, you may feel like throwing in the towel and give up attempting to classify a single disease based on 4 different, conflicting criteria. But just like when you really need that clutch 3-pointer to save the game, our specialist, DNAJB9 appears to save the day for fibrillary GN.
What exactly is DNAJB9? DNAJB9 is a specific immunohistochemical marker for fibrillary glomerulopathy. Its full name is DnaJ heat shock protein family (Hsp40) member B9 (DNAJB9). It inhibits the pro-apoptotic function of p53 and appears to be involved in the endoplasmic reticulum stress/unfolded response pathway. Despite locating this overabundant protein exclusively in FGN and not in other glomerular diseases or healthy individuals, the pathogenesis remains a mystery. However, it is a specific marker that is detected by routine immunohistochemistry. If you thought the numbers for PLA2R and membranous looked impressive, DNAJB9 has 97.6% sensitivity and 99.2% specificity for FGN. You read that right! It is almost the perfect test! And fortunately, no more patients with FGN will be misdiagnosed as amyloid or diabetes! Granted, the disease is much less common than MN, but this is an easy win.
Once we finally identify this disease, what do we do about it? Well, this is where the news is less promising. Fibrillary GN is usually a kidney-limited condition that may progress to ESKD in 44% of individuals within 4 years of diagnosis. The high risk of recurrence after transplantation is another concern (up to 36% of cases in a single-center study). Even though it has been described as idiopathic, some associated entities include malignancy (23%), dysproteinemias (17%), and autoimmune diseases (15%). Immunosuppressive therapy has not shown improved kidney survival, although small cohort studies have shown some potential promise with rituximab, with a third of patients responding. Given how rare this disease is, randomized trials in the future are a pipe dream, but the discovery of DNAJB9 is the first step to understanding the disease and may potentially unlock the key to a future treatment.
Immunoglobulin G4 (IgG4)-related disease is a true chameleon with multisystem involvement in more than 80% of the cases that usually affects middle-aged to older males. The pancreas was the first organ identified as a target of this autoimmune condition in association with elevated IgG4 levels. Other sites of involvement were later identified including the biliary tract, lacrimal and salivary glands, lymph nodes, lungs, and kidneys. Serologically, patients present with positive antinuclear antibodies (ANA) (31%), peripheral eosinophilia (33%), hypocomplementemia (56%), and high total IgG or IgG4 levels in the serum (79%). Imaging is characteristic for mass formation in the affected organs and due to its variable clinical presentation, the diagnosis can often elude even the most experienced clinicians.
The kidney manifestations are another reason this disease is a chameleon. In this series, 77% of the patients had some degree of kidney dysfunction at the time of biopsy, with tubulointerstitial nephritis (TIN) being the most frequent finding. Interestingly, 26% of the patients underwent a biopsy or nephrectomy because of a renal mass detected radiographically. Hydronephrosis with IgG4-related retroperitoneal fibrosis can also be another phenotype, and all of these presentations can be acute, subacute, or chronic. Other commonly observed laboratory findings are proteinuria (usually <0.5 g/d) and hematuria, but rare cases of nephrotic syndrome have also been described.
Considering that the most common manifestation of IgG4-related disease in the kidneys is TIN, let’s review the characteristic findings on biopsy. Saeki et al described the clinicopathological characteristics of this condition. Light microscopy classically shows a predominant plasma cell and lymphocyte infiltration affecting the interstitium in a diffuse pattern. Eosinophils and an occasional patchy infiltrate were also seen in some cases. The interstitial fibrosis was characterized by a swirling pattern of the collagen fibers arranged as “spokes of cartwheel” or a “bird’s eye”. This distinctive feature named “storiform fibrosis” is best seen with periodic acid–Schiff and Jones silver stains. The glomeruli typically are uninvolved, although co-existing MN, mild mesangioproliferative glomerulonephritis, and IgA nephropathy have been observed.
In IgG4-related kidney cases with pure TIN, more than 80% of the biopsies show a granular tubular basement membrane deposition for IgG, C3, and kappa and lambda on IF. Electron microscopy shows immune-type tubular basement membrane deposits correlating with the IF findings. However, our chameleonic disease shows a mixed pattern of TIN plus other glomerulonephritis in certain cases. The most common one is MN (with its typical subepithelial IgG deposits along the GBM, PLA2R negative) seen in 7%-10% of the cases, and with less frequency mesangioproliferative glomerulonephritis, and IgA nephropathy (mesangial IgA deposition).
As one may imagine, besides the LM findings of “storiform fibrosis”, it may be difficult to suspect that IgG4 disease is behind all this. Fortunately, immunostaining for IgG4 is a game-changer! The interstitial plasma cell infiltrate was IgG4-positive in a distinctive way (>10 IgG4-positive plasma cells per high-powered field, and/or 40% IgG4/IgG-positive plasma cells). Cases of IgG4-positive TIN also stained positive for IgG4 in the plasma-cell rich interstitium, the glomerular capillary walls, and even the tubule basement membranes. Sharma et al reported IgG4 related tubulointerstitial nephritis that also involved the arteries!
Undoubtedly this stain narrows the differential diagnosis, but IgG4-related disease is not the only condition with this feature. Despite its name, the specificity of the immunostain can also be seen in diseases such as ANCA-associated vasculitis, which may also have a large amount of IgG4 positive plasma cells and elevated IgG4 serum levels. Furthermore, 15% of TIN of other etiologies also show IgG4 positive plasma cells. These include idiopathic TIN, lupus nephritis, MN, and even our common diabetic nephropathy. So despite the 100% sensitivity and 92% specificity, there are still other diseases to be excluded when the immunostain comes back positive.
Since clinical and pathologic findings alone are insufficient for the diagnosis, criteria for IgG4 disease in each specific organ have been proposed. In 2011, a group in the United States and Japan proposed diagnostic criteria for IgG4 related kidney disease that include a combination of radiologic and laboratory findings (ie, elevated serum IgG4 levels ≥ 135 mg/dL), histopathologic features in kidney biopsy, and evidence of the disease in other organs.
The good news is that the initial treatment of IgG4-related kidney disease involves corticosteroids (prednisolone 0.6 mg/kg daily- usually 30 to 40 mg/d), and that it tends to be a very steroid-responsive disease. Typically, the disease improves within a month of steroid initiation, although the degree of recovery unsurprisingly depends on the severity of interstitial fibrosis. Importantly, relapse has been reported in approximately 20% of cases, and other agents have been used not only as steroid-sparing strategies but also as interventions to lower the potential for relapses in systemic disease. With a rare disease, data are sparse, but some of these additional therapies include cyclophosphamide, rituximab, and/or mycophenolate.
In this second matchup of rare disorders, both DNAJB9 and IgG4 immunostaining boast impressive stats, with high sensitivities and specificities that help guide diagnosis and therapy. Which stain will ultimately prevail in this hard-fought contest?