So just by way of introduction, Dr. Salant has been very active, and one of the premier physician scientists, a true role model. And all of you, I know, have taken care of patients with membranous nephropathy, the most common cause of idiopathic nephrotic syndrome in adults. David Salant made a landmark discovery in terms of discovering the antigen for membranous nephropathy, and so we're really pleased that he's here. 

By way of background, he is professor of medicine and vice chair for research at Boston University School of Medicine. He received his medical degree in clinical training in South Africa. And then he did his research training at Boston University with Dr. Bill Kauser and joined the faculty in 1979. 

He has really been interested in the immunology of glomerular diseases, and in his landmark publication in the New England Journal of Medicine in 2009, he along with Dr. Beck and others discovered the target antigen in membranous nephropathy. And ever since, I think, we've had a better understanding of the disease and potential treatment options. 

Dr. Salant is the past chairman of the ABIM Subspecialty Board of Examiners in Nephrology. And he's a recipient of many awards, including election to the American Society of Clinical Investigation, the Association of American Physicians. He's an established investigator from the American Heart Association. He won the John Peters Award from the American Sociey of Nephrology, the Jean Hamburger Award from the International Society of Nephrology, and the Don Seldin Award from the National Kidney Foundation. 

He was also the recipient of the Edward N. Gibbs Award and lectureship from the New York Academy of Sciences. Today, he's going to talk about membranous nephropathy from bench discovering a major antigen. Thank you, David. 

[APPLAUSE] 

Thanks, Bob. Can you all hear me at the back? Good. So having spent the past 24 hours listening to what's happening in the nephrology training program here, there's no way you people-- residents should not be going into nephrology and staying here. 

This is a spectacular training program. It really is. If I were starting off my training at this point hearing about what's going on here, even if I were not interested in becoming an investigator long-term, this is a spectacular program, and I encourage you to look carefully at it, those of you that haven't already made decisions for next year. But you won't go wrong. 

So I do have some disclosures. The one that's most relevant to this talk is the diagnostics for membranous nephropathy. And I also write a chapter in Up to Date on membranous nephropathy. The objectives of this presentation are listed over here, and I'm not going to go through them. 

But I thought I would start with a case of a patient that we encountered about 10 years ago. And he was a 60-year-old man who arrived in the emergency room with dyspnea. He had had no chest pain, cough, or fever, but his chest X-rays showed a large pleural effusion. A pleural tap revealed 500 mL of serosanguinous fluid, and a chest CTC TPI showed a large left pulmonary embolus. So he had an abdominal CAT scan that was unrevealing except for the presence of a renal vein thrombosis in the right renal vein. And there was no evidence of malignancy. 

So he was admitted, started on dalteparin and transitioned to Coumadin. And because his urine analysis showed 3+ albumin in a bit of blood, a renal consult was called. 

So he got a history of having had a thrombotic stroke three months earlier. His family history was negative for thrombotic events or renal disease. He had anasarca and normal blood pressure. His urine analysis showed lipiduria, red cells but no red cell casts. Serum albumin was really low at 2.5, but his serum creatinine was normal. He was hypolipidemic. Hemoglobin A1C was normal. 25-hydroxyvitamin D was really low. And his urine protein on a 24-hour urine was 6 grams. And during the stay in hospital, the urine protein creatinine ratio ranged from 4 to 7. 

Workup for various autoimmune diseases and hepatitis serology were negative. His complement levels were normal. And he had a negative workup for monoclonal gammopathy. 

And because of the thrombosis, he had an extensive workup for thrombophilia that was negative, so he was started on furosemide, an ACE inhibitor, statin and Coumadin, and discharged. 

And I'll come back to this case a little bit later. But at that time, no one was willing to stop this anticoagulation to do a kidney biopsy because of the high risk. But he clearly had nephrotic syndrome. 

Three cardinal features are albuminuria hypoalbuminemia, and edema. And some of the complications-- he had hyperlipidemia, thrombophilia, vitamin D deficiency. We don't know about osteomalacia in him, and that's something that tends to occur in patients with longstanding nephrotic syndrome, particularly children and adolescents. And he had not had any infection. There is an increased risk of gram positive infections, which he didn't have. 

And so our fellow who saw him went through the usual differential diagnosis of adult nephrotic syndrome that's listed here. And I'm not going to go through all of these. But most of them were ruled out. The secondary causes that are on the right-hand side of the slide were ruled out. And he probably didn't have minimal change disease, but could have had any one of the three things focal sclerosis, membranous nephropathy, or membranoproliferative glumeronephritis. 

So of course, we wouldn't be talking about this case if it didn't turn out at some point to have a diagnosis of membranous nephropathy. And so what is membranous nephropathy? It's a leading cause of nephrotic syndrome in adults. It's an antibody-mediated kidney disease. And most patients have what we call primary or idiopathic membranous nephropathy, which is an organ-specific autoimmune disease that follows a variable course, as shown on the slide. 

There are important secondary causes, including class 5 lupus nephritis, hepatitis B, various drugs, particularly nonsteroidals, and various cancers. And there are some interesting alloimmune causes of nephrotic syndrome. There's an interesting neonatal form that I might say a few words about. It can be seen de novo after kidney transplantation, and also after allogeneic stem cell transplantation. It has been seen in occasional patients with glycogen storage diseases that have been on enzyme replacement therapy. 

So we're all on the same page. Let's just look at the glomerulus. And let me be sure that I can find the pointer. [INAUDIBLE]. 

So here, we see a normal glomerulus, with a very high-powered view. And we see here the open capillary loops. And on the inside of the capillary loops, you really don't see the endothelial cells. 

But occasionally, you'll see an endothelial cell bulging into the lumen. The space between these capillary loops is made up by the mesangium and the mesangial cells. But on the outside are these cells called podocytes. And the cell bodies reside in the urinary space. 

And if we take a cross-section through one of those capillary loops, what we see are these interdigitating foot processors of the podocytes that send down these primary and secondary processes that then separate into these interdigitating foot processors, bridged by what we call the filtration slit diaphragm. And the podocytes are separated from these fenestrated endothelial cells by the glomerular basement membrane. So the filtration pathway is through the fenestrated endothelial cells across the basement membrane, and through these filtration slits with the final barrier to albumin permeation being the filtration slit diaphragm. 

So it turns out, as I'll show you, that the target of injury in membranous nephropathy is the podocyte. And what I'd ask you to do is imagine that you're snorkeling in the urinary space, looking down on the surface of a glomerulus, shown here by this wonderful scanning electron micrograph prepared by one of our graduate students. And what you see are these octopus-like cells sending down their tentacles. These primary and secondary processes that then break up into these interdigitating foot processes. And these podocytes are embracing the glomerular capillary loops. 

Now, in membranous nephropathy, what happens is that there are immune deposits that form in the subepithelial space below the foot processors. So here once again is the normal structure. And we see here what happens in membranous nephropathy. These electron-dense deposits form under the epithelial cell-- under the foot processors, and cause disruption of the normal architecture of the podocyte foot processes so that they become what we call effaced. 

And there is also disruption of the filtration slits. And that leads to altered permeability, allowing for protenuria And these immune deposits are formed by immunoglobulins and complement. And the immunoglobulin deposits that form are due to predominantly the IgG4 a subclass of IgG. 

So the standard therapy for idiopathic or primary membranous nephropathy is, as an outpatient, to start ACE inhibitors or ARBs or statin, diuretics if the patients are at risk for thrombotic events. And that's typically if they have very low [INAUDIBLE] levels-- typically below about three. Patients are often treated prophylactically with Coumadin. And if there is no resolution in three to six months, one or other immunosuppressive regimen is instituted. 

So what we learned about this disease dates back to almost, in fact, a little over 50 years, when a fascinating model was developed by Dr. Walter Heymann, who serendipitously immunized rats with a tubular brush border fraction from a normal rat kidney, and found that they developed protenuria urea after about six to eight weeks, and an immune complex glomerular nephritis that was indistinguishable from human membranous nephropathy. Now, to do any studies, to understand the pathogenesis of this disease, six to eight weeks is a long time to wait. So when I moved to Boston in the late 1970s to work with Bill Couser, we started working with what we call a passive model of Heymann nephritis. 

Passive Heymann nephritis is induced by immunizing a third-party animal-- that we chose to immunize sheep with this brush border fraction called Fx1A developed sheep antibodies, two of the antigens, in Fx1A injected them into rats. And they became protenuric within five days. Now, at that time, it was believed that these immune deposits that I showed you were formed by the entrapment of circulating immune complexes. 

But we had reason to believe that that might be an incorrect model, and did a study using an isolated perfused rat kidney system, in which we perfused the kidneys with the antibodies to a Fx1A so that they could not recirculate. So there's no way that immune complexes could form. And what we found, as in the human disease, as in passive Heymann nephritis, there are these granular immune deposits that formed in the glomerulus. 

And even within one hour of perfusion, we could begin to see these small subepithelial electron densities. So we stated at the time that our studies are consistent with our hypothesis that the antigen is a constituent of the glomerular capillary wall, and maybe a constituent of the epithelial cell surface. But we had no concrete evidence as to what that antigen might be. 

At the same time, we did some studies in the passive Heymann nephritis model in which we depleted the animals of complement. Because as I mentioned, there's a lot of complement deposited in the glomeruli of patients with membranous nephropathy. And lo and behold, what we found is if we depleted the animals of complement, they had no protenuria, as compared to the non-complement depleted animals. 

And in a subsequent study, Dr. Andrey Cybulsky who was a fellow in the lab, went back to the isolated perfused kidney-- perfused kidneys that already had the immune deposits with C8 or C6-deficient human serum got no albumin urea. But if he reconstituted the deficiency, got massive albumin urea, together with severe podocyte injury. And this was due to the formation of the membrane attack complex of complement, whereas in the kidneys perfused with C8-deficient serum, there was no injury, even though there were these immune deposits. 

So what was the antigen? Several years later, a group at Yale, Dr. Kerjaschki working with Marilyn Farquhar, actually identified the target antigen. And it took a few years for the antigen to be cloned. And it turns out to be a protein that's called megalin that's expressed on the soles of the foot processes of the podocytes in these coated pits. Now, megalin is also expressed in the proximal tubular brush border, which is why immunizing rats with proximal tubular brush border caused an immune reaction that led to antibodies that bound to megalin on the soles of the foot processes of the rat podocytes. 

Now, just to go back, it turns out that megalin, although many people, including ourselves, looked to see whether megalin might be the target antigen in human disease, human membranous nephropathy, megalin is not expressed on the podocytes in humans. And there were no circulating antibodies directed against megalin. So there was a lot of skepticism about whether or not this animal model was relevant to the human disease, until some years later in 2002, 

Dr. Pierre Ronco and his colleagues in Paris were faced with a baby that was born with membranous nephropathy. This was a baby with antenatal membranous nephropathy. And rather than just passing this off, they actually investigated this baby and the mother. And what they found was that the mother was homozygous-deficient in a protein called neutral endopeptidase, which they had previously shown in experimental studies to be expressed on podocytes. And the baby and the mother had circulating antibodies to neutral endopeptidase. 

And how was that? Well, it turns out that the mother had had a miscarriage about six months before. And the father of both the miscarried baby, as well as the neonate born with membranous nephropathy, was positive for neutral endopeptidase. So the first and miscarried baby immunized the mother to NEP. 

She then had these circulating antibodies that crossed the basement, that crossed the capillary wall, bound to NEP in utero and closed the neonatal membranous nephropathy. And so this proved that the pathogenesis of membranous nephropathy involves reactivity of circulating antibodies with antigens expressed on podocytes. However, NEP antibodies do not account for membranous nephropathy in adult humans. 

So is there another intrinsically glomerular antigen in idiopathic membranous nephropathy? So we embarked on a series of studies. And we've done this in the past, and been unsuccessful. And several other investigators had done some of this studies. And no one had ever really come up with an antigen. 

But we had had some success with some new proteomic techniques. So we decided to have another try. And so we get these kidneys that are discarded from the new donor kidneys, normal human donor kidneys, that can't be used for transplantation. We get these from the New England Organ Bank. 

We minced them up. We sieved them. We isolated the glomeruli. We homogenized the glomeruli. We extract the proteins, and we end up with this crude extract of human glomerular proteins. 

We then clear them, and then we run them out on polyacrylamide gels, transfer them to membranes, and do standard western blots. And we'd done this in the past, and other people had done some other things. And no one had ever come up with a consistent antigen. 

And then one day, Dr. Larry Beck, who was then a fellow in my lab, walked into my office with this western blot. And this was essentially the Eureka moment. Larry had Western blotted these glomerular extracts with serum from five patients with membranous neuropathy and five controls. Also protenuric, diabetics, focal sclerosis. And he only got this band in the patient's serum from patients with membranous nephropathy. 

So why did we get lucky. Well, the key to success here was that we happened to use non-reducing conditions. So let me explain. 

Normally, when one does a does a polyacrylamide gel electrophoresis, you use a reducing agent, because that unfolds the protein. Allows the protein to be nicely coated with SDS, and you get nice clean bands. But it alters the conformation of the protein. If you use non-reducing conditions, you retain the tertiary structure of the protein. And that's why we got lucky. 

And to prove that these proteins were all the same, we actually did something else. We chopped off the sugars from what we assume to be glycoproteins, and got a similar shift in the size of the band. So we were confident that these were all the same protein, and that the antibodies actually identified the backbone of the protein component of the antigen. 

But we didn't know what the antigen was. So we then cut out these bands. Sent them off to our collaborators in Louisville, Kentucky-- Dr. Jon Klein and his colleagues. And they sequenced the protein using mass spec, and sent us a list of about 26 different proteins. So we had to work our way through this list, because there are a lot of proteins at about 180 kilodaltons in glomeruli. 

And so the question was, what was the antigen? We went through this partial purification. We separated by gel electrophoresis 

We did the mass spec. We ended up with several potential proteins. And now we had to work our way through. 

And we presented the data at the American Society of Nephrology. And this was about a year later. And Larry Beck, who was doing this work as a postdoc fellow, was seen to be walking on the Golden Gate Bridge eyeing this sign, while I was waking up at 3:00 in the morning thinking I was destroying the career of this promising young physician scientist. 

Anyway, we eventually got to the bottom of the barrel. And there was one protein called the phospholipase A2 receptor that we had never heard of, and we'd ignored, because we didn't know, we had not heard about it. It was not known to be expressed on podocytes. And we went to the literature and found two groups that had been working on this in the past. 

One group in Japan didn't respond to our emails. And the other person, Gerard Lambeau, had originally cloned the gene for PLA2R. He's a person that had worked for many, many years on secreted phospholipase and using a ligand binding assay had discovered this receptor. And he had produced some antibodies and had the original constructs in his freezer, and hadn't worked on this protein for about 12 years. 

And he was very excited. And so he sent us some of these constructs so that we could express the recombinant protein. And he sent us some of his rabbit anti-PLA2R And so we ran the anti-PLA2R against the human glomerular extract, with and without deglycosylation, and got the same two bands that we have seen with the human serum. And that corresponded to PLA2R. 

And when we used membranous nephropathy serum against the recombinant PLA2R in the human glomerular extract, we saw the identical pattern. So we were confident that our patient membranous nephropathy sera was identifying PLA2R We did some additional confirmatory studies using immunoprecipitation, and blotting back to confirm this. 

And the other thing that we did, using Gerard's antibody, his rabbit antibody PLA2R, we were able to show that it's expressed where we expected it to be expressed, shown in red on the podocytes which lie outside the glomerular basement membrane, shown here in green, using an antibody to a matrix protein called agrin. The other thing that we found was that, when patients develop membranous nephropathy, the PLA2R that had been on the surface of the podocytes redistributes, and comes to co-localize with the IgG in the immune deposits of patients with primary membranous or idiopathic membranous nephropathy, but not in secondary membranous nephropathy, such as lupus nephritis. 

The other thing that Larry did was to actually elute the antibodies out of the glomeruli from patients with membranous nephropathy that had had kidney biopsies, and showed that the antibodies eluted from the glomeruli were reactive with PLA2R and the human native protein. And the circulating antibodies were reactive with PLA2R. These are Western blots run against PLA2R using the human serum-- six of them. And the predominant subclass of the antibodies was IgG4. 

So once we presented this, a number of people from around the world, around the country, started sending us banked sera. And so we had started for the relatively small number of cases from Boston. But we got samples from the Mayo Clinic, from the Netherlands, from Sweden, from China. And it turned out that, on Western blot, anywhere between 72 and 82 of patients with primary membranous nephropathy have circulating antibodies to PLA2R, whereas none of those were secondary membranous nephropathy in our hands. All the disease controls or normal controls was positive. 

So we were delighted when, about 18 months after our paper came out in the New England Journal, this paper appeared also in the New England Journal. And this was a genome-wide association study done by a consortium of investigators in Europe and the UK, in which they had accumulated cases of a fairly large number of patients with primary membranous nephropathy. And they had done this genome-wide association study, and come up with two hits. 

The one was an association with MHC class II, this HLA-DQA1, which was not very surprising, because membranous nephropathy is an autoimmune disease, and most autoimmune disease-- and in fact, all autoimmune diseases as we know them, are associated with the class II MHC. But there was a very strong association with DQA1. But remarkably, there was also an association with a single nucleotide polymorphism in the chronic region of PLA2R. 

And they didn't know what to make of this, because they had also never heard of PLAR2 until our paper came out. And they published [INAUDIBLE] showing these Manhattan plots showing a strong association with DQA1 and PLA2R. And the remarkable thing is that if patients are homozygous for both the slip in PLA2R and the HLA-DQA1, polymorphism there's almost an 80-fold odds ratio that they will get membranous nephropathy. 

So what about the MHC class II association? Well, this has been confirmed and linked to antibody levels and prognosis in caucasians and southeast Asians. But there is ethnic variation, because in China, rather than HLA-DQA1, the predominant association, which is with a different MHC class II DRB1 of 1501 0301-- and this is in Han Chinese, and a strong association linking this to the prevalence of anti-PLA2R-associated membranous nephropathy. And more recent modeling studies have identified amino acid variants within the peptide binding groove of the DR beta chain that are predicted to bind putative epitopes in PLA2R. 

So to summarize up to this point, membranous autoantibodies identify a confirmation-dependent epitope on PLA2R. 70% to 80% prevalence of circulating anti-PLA2R in primary membranous nephropathy. The antigen is expressed in podocytes. There's co-localization of PLA2R with IgG4 within the immune deposits. One can elute the antibodies from primary members from biopsy samples. And there's a strong genetic association of PLA2R variants and MHC class II with primary membranous nephropathy. 

So going back to our case. So what about that patient? No one was willing to do a kidney biopsy on him. And he arrived in the outpatient clinic three months after his presentation with a pulmonary embolus. 

He had persistent anasarca now requiring more diuretics. His GFR remained normal. And the question was, should he be admitted for a kidney biopsy, which would have meant holding his Coumadin, bridging him with heparin. Should he be treated empirically for what, or were the other options? 

And the other option was to do the immunoassay. It turned out that he was strongly positive for antibodies PLA2R as compared to the normal control. So he was treated for membranous nephropathy initially with cyclophosphamide and prednisone. Didn't respond to that. And a few months later, received rituximab, and eventually went into remission. 

So now we have three possible immunoassays-- in fact, four for anti-PLA2R. The western blots are really not suitable for clinical use because they're labor-intensive and expensive. There is a commercially immunoassay based on an immunofluorescence test-- little bit like doing an anti-nucleide antibody. And there is also a specific and sensitive ELISA assay. And now the pathologists can also stain the kidney biopsies for PMA2R, and help differentiate primary from secondary membranous nephropathy. 

And in fact, we were asked to do a comparative study looking at the different immunoassays in primary and secondary membranous nephropathy in preparation for a submission to the FDA. And we found once again that in the Western blot, 84% of these 95 cases was positive for anti-PLA2R antibodies. Similar number for the immunofluorescence assay. Less sensitive for the ELISA, and highly specific because none of the secondaries was positive, at least amongst this group that we examined. 

In July 2014, the FDA approved the commercial assay. And so the question that you might all have is, does the immunoassay correlate with disease activity? And so one of the early studies that we did was on banked samples from a group in the Netherlands. They had been following patients with membranous nephropathy for some time, and wisely had obtained serum samples that they had stored when the patients were nephrotic when they treated them, and when they relapsed. 

And they sent us these samples with blinded codes. And Larry Beck ran them out on Western blot, sent them back the results. They decoded them. And we found that when the patients were nephrotic, they had high levels of the antibodies. 

Most of those that went into remission had a resolution of the antibodies, and those that relapsed had a recurrence of the antibodies. And then Julia Hofstra who from the Netherlands also did an analysis comparing the titer of the antibodies to the amount of protenuria. And she found a fairly good correlation between antibody levels and the amount of protenuria. 

Another study done by a group in the UK, in Manchester, looked at the effect of antibody titer on the outcome of patients with membranous nephropathy. And they found that patients with high titers were more likely to end up on dialysis. This is dialysis-free survival. And so low and mid-range titers patients tended to preserve renal function, whereas those with high titers, about 50 of them actually went into end-stage kidney disease. 

Another study done in Italy, in Bergamo, on 81 patients treated with rituximab, they found that low titers in complete depletion of the antibodies six months after treatment with rituximab predicted complete remission. And all 25 complete remissions were preceded by complete antibody depletion. And reappearance of the antibodies predicted clinical relapses. 

In a collaborative study that we did with the group at the Mayo Clinic, they had also treated patients with rituximab several years before banked samples. And they sent us these samples. And once again, Larry Beck ran out the Western blots. And here, we see an example of one patient. He has the decline of the antibody levels over the course of two years after treatment with rituximab, whereas the protenuria lagged behind. 

And when the data was aggregated for all the cases, once again, the antibody levels after rituximab fell more rapidly than the protenuria. And that's important, because if you're treating a patient with immunosuppressive drugs, what do you follow? Do you follow the protenuria? 

And we argue now that following protenuria is the wrong thing to do, because the resolution of protenuria lags way behind the resolution of the immunological response. And here, we see that some patients will go into complete remission. But many patients will be left with some protenuria. And that's due to the remodeling and abnormal glomerular capillary wall that I showed you early one. 

So the question that arises now whenever I give this talk is, do you need a kidney biopsy if PLA2R is positive in a patient with nephrotic syndrome. I'll give you two examples. This was a 60-year-old man with a pulmonary embolism, stroke, from a paradoxical embolus from a renal vein thrombosis. 

Urine protein was 6 grams. Serum albumin was 2.5. And his anti-PLA2R was strongly positive by Western blot. This is like the case that I presented to you. And most people would argue that it would be riskier to biopsy this patient than to treat him, given the high specificity of the test. 

The second was a more recent case. This was a 72-year-old male transplant kidney donor. He had given a kidney to his son-in-law about 12 years before, and he presented with anasarca, urine protein of 12 grams a day, and serum albumin of 1.8. And the community nephrologist was reluctant to do a kidney biopsy on his single kidney because she had had some complications, and was with her interventional radiologists. And so she sent the patient to us. 

We ran his immunoassay, and was strongly positive by ELISA. And we argued that there was no need for a kidney biopsy. He, too, was treated with rituximab, and went into remission. 

So a fairly large study was done by a group in Hamburg, where they wanted to look at the sensitivity and specificity of the combination of the anti-PLA2R [INAUDIBLE] assay and the immunostaining for PLA2R on the kidney biopsy. And they had 73 patients with primary member nephropathy, and 15 secondary. And they found that the positive predictive value of a positive anti PLA2R or positive staining on the biopsy was 100%, and with a sensitivity of 84%, which goes along with all of the other studies that we've done pretty much from around the world-- except perhaps Japan, where it turns out that if you have a patient in whom you've excluded secondary causes of membranous nephropathy, you do the test. 

If the test is positive, that patient has primary membranous nephropathy. If it's negative, then they could still have primary membranous nephropathy, but not due to PLA2R. And we'll look into that. 

So one last thing before I move forward on the PLA2R-negative cases is, what about this issue of do you need to screen every patient with a membranous nephropathy for cancer, because of the well-known association of solid tumors and membranous nephropathy? So again, in a collaborative study with a group in the Netherlands, they had banked serum and a very good follow-up history on their patients with membranous nephropathy. And together, we found that if patients were positive for anti-PRA2R, it was very unlikely that they had a concomitant malignancy. 

That's not to say they wouldn't develop cancer later on. After all, these are predominantly middle-aged to elderly patients, particularly men. Whereas if they were negative anti-PLA2R, there was a much stronger chance that they might have cancer associated membranous nephropathy. And these are the patients, these PLA2R-negative patients, that merit a more careful look for an occult cancer. 

So one other thing that's emerged recently is this issue of epitope spreading. And so all patients will have antibodies to this region of the phospholipase A2 receptor. But it turns out that some patients will also develop antibodies that identify other parts of the protein. A study done by our collaborator, Gerard Lambeau in Nice, essentially found that if there is a spreading to other regions of the molecule, there's a greater chance that the patients will not undergo spontaneous remission, are no more likely to end up with kidney failure and requiring dialysis or a kidney transplant. Whereas those in whom the antibodies were restricted to that N terminal epitope, those are the ones that tend to do best. 

So just to end off, the question is, could there be another antigen? What about those that are anti-PLA2R-negative. And it turns out that very early on, when we were doing our studies, Larry Beck noticed that there were some patients that were negative for antibodies to PLA2R, but did have reactivity to the native antigen in human glomerular extract. And we wondered whether there was a novel antigen. And it also turned out to be IgG4 the antibodies also turned out to be IgG4. 

So what is the unknown antigen? Is it a variant of PLA2R with a unique configuration, or is it something completely different? So once again, we cut out this band, sent it off to our collaborators in Louisville, Kentucky. They did the mass spec, and sent us back possible proteins. And it turned out to be a protein called thrombospondin type 1 domain containing 7A 

So there's a bit of a background story to this. So if I can go back, when we were at this stage and we had just identified what we thought was a protein, Dr. Rolf Stahl in Hamburg was visiting Boston with his wife. We went out to dinner, and he said, you know, David. I think we've identified another antigen in the PLA2R-negative cases. 

So I smiled. And he said, why are you smiling? So I said, well, I think Larry Beck has also identified another antigen in the PLA2R-negative cases. And his jaw dropped. And he went back and called Gerard Lambeau in Nice. 

Turns out that Rolf Stahl had sent a graduate student to work with Gerard Lambeau, and said, I want you to do exactly what Larry Beck did to identify PLA2R. And so he had done the same thing, and come up with the same band. And so Gerard Lambeau called and said, we need to get on the phone together. 

So there was this conference call-- Rolf Stahl in Hamburg, Nicola Tomas, the graduate student, and Gerard Lambeau in Nice, and Larry Beck and me and Boston. And we'll all beating around the bush. No one's really saying anything specific. And Nicola Tomas the graduate student, pops up and says, well, ours begins with T. 

[LAUGHTER] 

And Larry said, yeah. So does ours. And so everyone agreed at that point, there was no point in us competing. 

We got together. Nicola Tomas the graduate student, was the first author on that. And it's in a combination paper. This was published in the New England Journal of Medicine, showing that in some patients with PLA2R-negative membranous nephropathy with these antibodies THSD7A, similar co-localization of the IgG antibodies with the antigen. 

And for the sake of time, I think I'll end at this point, and summarize what I believe are the clinical advances. So we now know that anti-PLA2R is highly specific, and about 70% to 80% sensitive for the diagnosis of membranous nephropathy. The likelihood of a secondary cause, including malignancy, is low. If anti-PLA2R or tissue staining is positive, the presence, titer, and epitope spreading of anti-PLA2R reactivity correlates quite well with disease activity. 

About 10% of patients that test negative for anti-PLA2R have anti THSD7A, which may associate with cancer. I haven't shown you that, but there is some association. And the presence of anti-PLA2R prior to transplantation is associated with a high risk of recurrent disease. 

This is something I've also not shared with you, but turns out to be true in our hands and others. And anti-PLA2R and PLA2R staining distinguish recurrent and novo post-transplant membranous nephropathy. The transplantists may be interested in that. 

So I'll end at that point, and acknowledge the many people in Boston, Louisville, Hamburg, Nice, and the Mayo Clinic, the New England Organ Bank. We couldn't have done these studies without their cooperation. And of course, the families of the deceased kidney donors, and the patients and volunteers that provided serum. Thank you for your attention. 

[APPLAUSE] 

We have about four or five minute for questions. We're just waiting for a microphone. 

I was quite excited with your very interesting story, and I would say the outcome of [INAUDIBLE] and from this morning's talk. And now it looks as though PLA2R [INAUDIBLE] is really the case. Now, the question I have, which has intrigued me since this morning, is there something peculiar about the B cells in these patients? Because, I mean, usually with most autoimmune diseases, they make IgG bodies of different isotopes. What is peculiar about [INAUDIBLE] make IgG4? Has anyone looked into it? 

So that's an excellent question. So IgG4 is often regarded as a protective antibody. Beekeepers, for example, after they've been stung multiple times, develop IgG4 antibodies, which makes them resistant to further bee stings. However, there are a bunch of other autoimmune diseases that are IgG4-associated, including one order in uniform of myasthenia gravis, and blistering skin diseases, [INAUDIBLE]. 

So I don't know why these particular autoimmune diseases are uniquely associated with IgG4. And I'm not even absolutely certain that the IgG4 antibodies that we see are the pathogenic ones, because of course, we're seeing the patients well into to the autoimmune response. And it's possible that they may have had different isoforms of IgG earlier on in their disease. And as I pointed out in the one slide, they all tend to have some reactivity to IgG1 and IgG3, but much lower levels typically. So I'm not excluding the possibility that the IgG4 is a major appearance. 

And then the question is, if it is IgG4, how is it causing damage? Is it activating complements? Well, it's supposed to activate complement. IgG4 can't activate the classical pathway of complement. We have some data to suggest it may be able to activate elected pathway. 

So that might explain the presence of complement. But I don't have an answer to your question, but it's been a question in a number of other autoimmune diseases. And I have an interesting review that I can share with you, where someone is really pondering this and looking into these autoimmune diseases. 

[INAUDIBLE] thank you for the outstanding talk. There is recent data showing a potential link between pollution and membranous nephropathy in China. How is this pathway linked to that? 

Yes, so I actually talked about this in this morning's talk. So just to-- in certain regions of China, there's a very, very high level of air pollution. And it turns out that that maps very closely with remarkable increased incidence of primary membranous nephropathy in those regions. 

So we have a theory. PLA2R is also expressed in the lung at lower levels. And it's conceivable that with air pollution or smoking, or other infections, that PLA2R may be up-regulated expressed. And that may be the site for triggering the immune response. It's a hypothesis at the moment, unproven. 

Any other questions? Mine was a quick naive question. I apologize. But can you connect this with any of the NSAID nephropathy, phospholipids? Story? 

NSAID no? 

No, because-- well let me let me step back. I shouldn't say that so quickly. I think NSAIDs is a trigger. I don't think NSAIDs is the primary cause of membranous nephropathy. 

Sometimes, we see patients with PLA2R-associated membranous nephropathy that had had something else going on, and have taken NSAIDs, and then developed their protenuria and edema. So was it the preceding event that caused them to take NSAIDs, or was it the NSAIDs? I don't know. 

I just wondered if your test might predict the risk for NSAID [INAUDIBLE] upregulation of phospholipids perhaps. 

Genetically? 

Genetically or pharmacologically. 

It would have to be, because the antibodies become positive when the patient develops the disease. But your point is well taken. 

Thank you for a wonderful talk. I wondered if there is ever a natural variant in the human population where either of the antigens are missing. Because we still know what their function is, so if there's a natural, and more like Alports when you transplanted them, when you then make an alloimmune antibody to the antigen, and what would that look like? 

Yeah, as far as I know, there are no examples of that. Because if there's no antigen on the podocytes, and they were to develop the antibodies, how would we ever know? 

Well again, if you had a transplant scenario, where you have an amino antigen transplanted, then you'd have allo. And you could identify the antigen, because you'd have this disease. 

I hear you. What one would have to do is find all those cases that we are calling a recurrence and go back-- and we've done this with some of the cases, to see whether they did have PLA2R-associated primary membranous nephropathy as the cause of their primary disease. And if not, what did they have? That's a good thought. 

Thank you. 

Dr. Salant, thank you so much for being here. 

My pleasure. 

[APPLAUSE]