Expires October 2019
Neuromyelitis optica (NMO) is a severe, relapsing, autoimmune, inflammatory and demyelinating central nervous system disease that predominantly affects optic nerves and the spinal cord. The disorder is now recognized as a spectrum of autoimmunity (termed NMO spectrum disorders or NMOSD) targeting the astrocytic water channel aquaporin-4 (AQP4). Brain lesions are observed in 60% of patients with NMOSD and approximately 10% will be MS-like. It is estimated that half a million people are living with NMOSD in the world.
Many patients with NMOSD are misdiagnosed as having multiple sclerosis (MS). Importantly, the prognosis and optimal treatments for the 2 diseases differ. Beta-interferon, a treatment promoted for MS, exacerbates NMOSD. Therefore, early diagnosis and initiation of NMO-appropriate immunosuppressant treatment is important to optimize the clinical outcome by preventing further attacks and reducing disability.
Sean Pittock, M.D., is a Professor in the Department of Neurology, Director of the Clinical Neuroimmunology Laboratory, and Consultant in the Division of Clinical Biochemistry and Immunology at Mayo Clinic in Rochester, Minnesota.
Welcome to Mayo Medical Laboratory’s Hot Topics. These presentations provide short discussion of current topics and may be helpful to you in your practice.
Our speaker for this program is Dr. Sean Pittock, Professor in the Department of Neurology, Director of the Clinical Neuroimmunology Laboratory, as well as consultant in the Division of Clinical Biochemistry and Immunology at Mayo Clinic in Rochester, Minnesota. Dr. Pittock introduces a new FACS Assay for detection of Aquaporin4-IgG discovered in that lab, which is required for the diagnosis of neuromyelitis optica spectrum disorders and allows for distinction from multiple sclerosis. Thank you, Dr. Pittock, for presenting with us today.
Hi, my name is Sean Pittock and it’s a great privilege to be able to tell you about a very exciting new test that’s going to be made available through Mayo Medical Laboratories and through the Neuro-Immunology Laboratory. And, this is really the first time that a FACS assay has been used for the detection of auto antibodies, so this is a very exciting first for our laboratory.
So the test name is AQP4-IgG FACS assay. It’s going to be available for testing both serum and spinal fluid.
These are my disclosures.
So, first of all, before we start this presentation, we need to ask the following questions: How is this testing going to be used in your practice? When should the tests be used? And how will the results impact patient management?
So, first of all, the outline of my talk is to give you an overview of what neuromyelitis opticus spectrum disorders are. I’m then going to talk a little bit about the different assays that have been used to detect the water channel antibody and show you that this is an assay that is most sensitive and specific for the detection of this antibody. And then I’m going to introduce you to the actual FACS assay itself and show you how we actually interpret the findings and how we decide on whether a patient is positive or negative. And then I’m going to talk to you about what the optimal specimen type is for testing.
So, first of all, aquaporin4 antibodies are the first-ever neural antibody to be shown to be both sensitive and specific for any form of inflammatory demyelinating central nervous system disease. It’s interesting because the target, the water channel, is actually on an astrocyte and, traditionally, people investigating inflammatory demyelinating disorders were generally focused on the oligodendrocyte. So this discovery really resulted in somewhat of a seismic shift in the way we think about these diseases. Here we see a slide showing the immunostaining pattern of NMO IgG, which is targeting the water channel. Here, we’re looking at a patient’s serum binding to the aquaporin4 on mouse tissues. And here you can see mouse cerebellum, mouse midbrain, as well as the smooth muscle and mucosa, which are essentially negative. It was the discovery that this binding pattern was present in approximately 70% of patients with the disease that led to the recognition that this was a specific binding pattern, that these patients had a specific IgG, and it was subsequently work by Dr. Vanda Lennon in our laboratory that resulted in the identification of the target antigen as being the water channel.
NNO/AQP4-IgG Antibody (IF)
The water channel sits on astrocytes in the brain. Its highest concentration is in polarized plasma membrane, and here you can see it sits within a dystroglycan complex. It is present at the CNS perivascular and peripial end-feet that are in direct contact with the basal lamina of the endothelium and the pia mater. It’s also present in the ependymal cells and the subependymal layers lining the ventricles, which is important because we now know that lesions in the brain tend to localize in these areas, areas where there’s high expression of water channel. They are also present in hypothalamus and in interneuronal synaptic junctions. They are not found in neurons, oligodendrocytes, or choroidal epithelial cells.
NMO-IgG Targets AQP4- Where is the Brain is AQP4?
The discovery of the antibody really resulted in a change in the way we think about this disease, and we began to notice that patients that did not have the full spectrum of the disease in other words, did not fulfill the 1999 Wingerchuk criteria for NMO, that some patients who had inaugural or limited forms of the disease were also positive for the antibody. We began to recognize that a patient might have 1 episode of optic neuritis, be tested for the antibody being positive, or have 1 episode of longitudinal extensor transverse myelitis, be tested for the antibody and be positive; and that positivity for the antibody predicted relapsing disease, but not only that, indicated that these patients, in fact, had the same disease, an NMO spectrum disorder. It has now become evident that patients commonly have brain lesions prior to the discovery of the antibody. It was felt that, generally, patients with this disease did not have brain lesions or that their brain was spared. And we now know that 60% of patients will have brain lesions. So here you can see that the discovery of the antibody has resulted in a widening or broadening of the spectrum of clinical and radiological features associated with this disease, and that’s why we now refer to patients that are aquaporin4 antibody-positive that have these symptoms and signs as having an NMO spectrum disorder.
NMO Spectrum Defired by AQP4-IgG
So now we think of the disease in a much broader sense, and this diagram, hopefully, will help you, in some ways, visualize what we really are talking about. Patients with recurrent optic neuritis, for example, or rON, about 20% of those patients will be aquaporin4 antibody-positive. In other words, they have an NMO spectrum disorder. The other patients will have some other disease, perhaps multiple sclerosis or another demyelinating disease, or they will have some indeterminate diagnosis. But the 20% of patients that are positive for the water channel antibody, those patients have an NMO-spectrum disorder and should be considered autoimmune aquaporin4 channelopathy. Similarly, patients can have just relapsing LETM or a single episode of LETM, or they could have intractable nausea and vomiting, evidence of an area postrema involvement. Either way, these patients would fulfill the criteria for NMO-spectrum disorder with aquaporin4 antibodies.
“NMOSD” Replaced by “Autoimmune AQP4 Channelopathy”
Now, it’s very important to make this diagnosis because the identification of the antibody indicates a specific immunopathologically distinct disease. Furthermore, this disease is distinct for multiple sclerosis. Patients with multiple sclerosis do not have aquaporin4 antibodies. In fact, there is no biomarker specific for multiple sclerosis. Unfortunately, many patients with NMO or NMO spectrum disorders are often misdiagnosed as having MS and area treated with MS drugs; but many of the MS drugs make NMO or NMO spectrum disorders worse. So it is extremely critical that patients get the right diagnosis and, therefore, get the appropriate treatment. Furthermore, early diagnosis is important because patients with NMO or NMO spectrum disease that are not treated can develop significant disability. So a test that’s highly sensitive, that’s highly specific is extremely important. So which test is best? Well, as I said, we originally started off back in the late 90s with an indirect immunofluorescence assay as the method of detection. We then moved to immunoprecipitation assay or IPA. We then replaced that with an ELISA. And more recently, we have been offering a cell-based assay. In 2012, we worked with our colleagues and collaborators at Oxford to try and figure, in a blinded way, which assay was best. And so we sent samples to a third party in Canada who blinded the samples and then sent half of each sample to Oxford Laboratories and also to Mayo Clinic Laboratories. And what we found is very, very interesting. Here you can see, in patients with NMO or NMO spectrum disease or high-risk NMO diseases, how frequently patients were positive by the different assays. And you can clearly see that the FACS assay on the left or the cell-based assay, CBAs, appear to perform better than the ELISA, or the immunoprecipitation assay, or the indirect immunofluorescence assay. So this was a very nice study because it was done in collaboration with 2 labs who were currently offering these types of tests in a blinded fashion, and I think it was very, very informative.
Which Test is Best?
It’s important to note that the FACS assay that was used in this study was actually used in the M23 isoform of aquaporin4, not the M1 isoform that we use, and I’ll try and explain to you, as we move forward through this talk, why that is. But first of all, just look here. You can see that the sensitivity of the original assay we used in the discovery of this antibody and in the subsequent detection of these antibodies until it was replaced by a more optimal assay, was only 48.3%, whereas the FACS assay had a 76.7% sensitivity. The specificities were very high here because these were highly selected patients.
Sensitivity and Specificity of 6 Aquaporin-4-IgG Assays
Now it’s important, at this point, to introduce the fact that aquaporin4 exists in 2 isoforms, the M23 isoform and the M1 isoform. Here you can see freeze-fracture analysis, and what you can note is that the M23 isoform exists as particles or higher order arrays on the surface of the astrocyte, whereas M1 tends not to form these higher order arrays. And we think this might be important in respect to specificity of the assay.
Characteristics of AQ
We cloned cDNA’s for both isoforms into expression vectors and produced HEK-293 clonal cell lines that express M1 or M23 AQP4. As shown in these freeze fracture images, the observed structures, circled in blue, on the surface are very different. With M23, there are large orthogonal arrays and in M1 the arrays are absent. This is an important point to understand and it relates directly to pathogenic mechanisms of NMO. The distance between tetramers in an array is perfectly suited for binding both F ab regions of an antibody, which is necessary for high affinity binding. The formation of arrays makes a perfect target for complement activating antibodies and subsequent major damage to a cell expressing these arrays.
What Isoform Provides Highest Clinical Specificity
In this Mayo Clinic study, first authored by Jim Fryer in our group, we looked at comparing the sensitivity and specificity of FACS assays using M1 isoform vs M23 isoform. We also compared with the ELISA assay, which uses the M1 isoform, and the cell-based assay, which is currently used in our laboratory, which also uses the M1 isoform. What we see is that generally the FACS performs better, certainly, than the ELISA, but these are low numbers of cases. However, what’s most important, though, is if we look at the controls. Some of these were healthy controls. Others were disease controls. But what you can see is that when using M23 FACS, you lose specificity. We dropped from a specificity of 100% to a specificity of 95% when using the M23 FACS. We suspect that this may be due to the fact that M23 is expressed as higher order arrays on the outside of the cell, and these may be more sticky or result in some nonspecific binding.
Seroprevalence of Aquaporin-4–IgG in a Northern California Population Representative Cohort of Multiple Sclerosis
We also need to think about specificity when it comes to comparison of our new FACS assay or cell-based assays with some of the other assays, for example, ELISA. This is a study where, in collaboration with Lisa Barcellos, we tested approximately 1,000 patients who were in her Northern Californian population representative cohort of multiple sclerosis. The patients were tested by 3 assays: the ELISA, the cell-based assay, and our M1 FACS assay. And what you can see here is that 7 patients were positive by ELISA, 3 patients were positive by cell-based assay, and 2 of the 1,000 patients were positive by the FACS assay. However, when we looked at the clinical records of these patients, we confirmed that only 2 of the patients, specifically the 2 patients positive by FACS assay, had, in fact, the disease NMO or NMO spectrum disorder. Five of the patients that were positive by ELISA, in addition to another patient positive by cell-based assay, were actually false-positive results. This indicates to us that the FACS assay really performs best when we talk about specificity. And it’s very important to have a specific test, because if a patient has an episode of demyelination, or optic neuritis, or some clinically isolated event, and they’re positive for the antibody, they’re going to go on maintenance immunosuppressant medication. So a false-positive result potentially could result in an individual being put on life-long immunosuppressant therapy.
Introduction to the AQP4-IgG FACS Assay
Again, I’d just like to show you, this is a FACS assay. These are the results. Here I’d like to demonstrate how one can have a false-positive test result on a FACS assay. So the top group of FACSs are using the M23 isoform, and the bottom group are using the M1 isoform. In the left panel, you can see an NMO spectrum disease patient that is positive on both the M23 FACS and the M1 FACS. In the middle, you can see a patient that does not have NMO spectrum disease. This patient is positive on the M23 FACS, which is a false-positive, but is negative, appropriately, using the M1 FACS. On the right panel, you can see patients that don’t have NMO spectrum disease that are negative on both FACS assays.
AQP4-IgG evaluation: FACS3
How do we define whether or not a patient is positive or negative? It’s important to note that what we’re really comparing here is the signal that’s given off by the patient’s serum bound to a fluorescent antibody, when it binds to cells that are transfected with the water channel, compared to the signal that’s given off by the patient’s serum binding, or not, to cells that are not transfected with the water channel. What we have done is we have devised a ratio whereby we divide the mean fluorescence intensity of the GFP aquaporin4 antibody-positive population by the mean fluorescence intensity of the nontransfected cells. And, by testing large numbers of patients as well as controls, we have defined positivity as a binding index of greater than or equal to 2 as being positive. As you can see on the left, this is the readout that we get from a negative patient or binding to the nontransfected cells; whereas on the right, we see the positivity, the binding of the patient’s antibodies to the GFP-linked aquaporin4 in the transfected cells.
Should Serum, Cerebrospinal Fluid, or Both be Submitted for Testing?
We screen serum at a dilution of 1 in 5, and we screen CSF at a dilution of 1 in 2. If the binding indices are greater than or equal to 2, then it’s a positive result. If they are less than 2, then it’s a negative result. If the patient is positive, then we will titrate out to see what the end dilution is. So, for example, in serum, if a patient is positive at 1 in 5, then we will perform the testing at a 1 in 10, 1 in 100, 1 in 1,000, and so on dilution out to 100,000. And the point at which the binding index drops to less than 2 is considered the end dilution. In other words, the last dilution that is greater than 2 is considered the endpoint. Most patients’ titers generally fall around the 5, 10, 100, or 1,000 mark. Generally, titers do not go beyond 100,000, and the same approach is taken for spinal fluid. Thus, with the FACS assay, now we will be able to offer you an end dilution or an end titer for both these assays.
It’s important to note, however, that we currently do not know whether or not titer is clinically relevant. We certainly do see titers commonly drop in the setting of immunosuppression, and some data suggests the titers of antibodies will rise before an attack; but this really requires confirmation with some larger studies that are currently underway at Mayo Clinic.
The next question that I’m hoping to answer is: What specimens should be submitted? Should you submit serum? Should you submit cerebrospinal fluid? Or should you submit both for optimal testing for aquaporin4 antibodies? I think this is important because there has been an increased drive to try and get clinicians to submit spinal fluid for evaluation of encephalopathy. For example, we now know that NMDA receptor antibodies are probably more sensitively detected in CSF than in serum. What you can see here is that over the past 8 years, the frequency of CSF submission as the initial test specimen for aquaporin4-IgG testing has increased 10-fold. In other words, in 2007, 2.37% of all specimens submitted for aquaporin4 antibody testing were CSF. This year, approximately 20%, or 1 in 5 specimens, are spinal fluid.
This is a busy slide but really just look at the colors, and what you can see is that if you take NMO, these are patients that fulfill NMO criteria in yellow, you can see when we looked at paired serum and CSF examples in other words, this is where the serum and the CSF have been drawn within a certain short timeframe of each other, you can see that the serum FACS is the most sensitive, followed by the serum cell-based assay. But the CSF FACS or CSF cell-based assay detects less than the serum in all cases. Thus, this suggests that it is possible that if you only submit the CSF that you are actually potentially going to miss a positive result.
Should Serum, Cerebrospinal Fluid, or Both be Submitted for Testing?
This is further illustrated here in this figure. Here you can see that on the left panel, these are patients that are positive in serum, but are negative in spinal fluid. And what you can note is that in all of those cases, the titer in the serum is low. In the right panel, you can see these are patients that are positive in serum, as well as being positive in spinal fluid, and their serum titers are generally higher, in most cases greater than or equal to 100. This indicates that the production of aquaporin4-IgG occurs predominantly in peripheral lymphoid tissues rather than intrathecal production, and that a critical serum CSF gradient is required for the immunoglobulin, the aquaporin4-IgG, to penetrate the central nervous system. These studies indicate that the optimal specimen for testing for aquaporin4-IgG by FACS or cell-based assay is serum. And, if you suspect a patient has NMO spectrum disease, we strongly recommend that you submit serum and not spinal fluid.
Take Home Points
So we’ve covered a lot. We have discussed with you what NMO spectrum disease is. We’ve talked about the discovery of the aquaporin4 antibody, how that discovery has led to an expanding and broadening spectrum of the disease. We’ve talked to you about the importance of detecting aquaporin4 antibodies and how detection of the antibodies allows you to confirm the diagnosis, but also not only allows early diagnosis but allows early initiation of attack prevention therapies. We’ve talked to you about what this new FACS assay does, what a positive result is, how we titrate a positive result to provide you with an end-dilution or an end-titer, and we’ve finished up by telling you the importance of submission of serum, rather than spinal fluid, as the optimal and most cost-effective specimen for aquaporin4-IgG testing.
Thank you very much.