Behind the scenes in neural antibody testing

Expires: December 14, 2023

  • Presenter

    John Mills, Ph.D.
    Co-director, Neuroimmunology Laboratory
    Assistant Professor of Laboratory Medicine and Pathology
    Mayo Clinic, Rochester, Minnesota

    Questions?

    Contact us: mcleducation@mayo.edu.

    Transcript and references

     

    Introduction

    Hello, I'm John Mills. I'm the co-director of the Neuroimmunology Lab, and I also serve as a clinical consultant in the Genomics Laboratory here at Mayo Clinic. 

    Disclosures

    I have no disclosures for this talk. 

    Learning objectives

    Our learning objectives today are first to understand the central role of tissue amino fluorescence in the identification of neural antibodies; to recognize the impact of interpreter reader training in the quality of neural antibody evaluations; and to appreciate the benefits of an integrated laboratory approach to the development and validation of novel antibody biomarkers. 

    Tissue indirect immunofluorescence (IIF)

    I'm going to start off by talking about tissue indirect immunofluorescence. In testing for neural antibodies, tissue immunofluorescence is the central element for testing for these antibodies. It serves as the most comprehensive multi-analyte screening platform available for neural antibodies, provides a high degree of specificity when interpreted properly, and allows novel antibodies (neural specific antibodies) to be characterized and discovered. 

    Unfortunately, not all neural antibodies are optimally detected using tissue immunofluorescence due to limited sensitivity for some neural antibodies, an example, the LGI1 and Caspr2 antibodies. And for that reason, non-tissue based antigen-specific screening assays, such as cell-based assays, are also incorporated into evaluations and testing algorithms. It's imperative that these assays provide a high degree of clinical specificity. For some samples, staining patterns may be ambiguous even to the trained eye. And therefore we utilize confirmatory methods, such as line blots, to aid in the final interpretation. We avoid using these methods in isolation, as they may lack specificity in the absence of an appropriate tissue-based staining pattern. 

    For more technical detail on the right, I'm showing the standard process that we follow for tissue amino fluorescents. We utilize a custom-made section of brain tissue and accompanying non-brain tissue as a substrate. This tissue is fixed, blocked, and subsequently is incubated with the patients sera/CSF. After washing, an anti-human secondary antibody that is conjugated to a fluorophore is added to detect bound antibodies. The final, and by far the most critical, step is the interpretation of the staining patterns. This is critical. Firstly, because the staining patterns observed are complex, and the fact that this central testing drives downstream testing and often plays a critical role in the final interpretation. 

    Tissue IIF review process

    Interpretation of tissue immunofluorescence is by far the most critical step of testing and can make or break the quality of a neural antibody evaluation. The complexity of the interpretation can be reduced by the use of non-brain tissues, which can aid in the identification of non-organ specific antibodies, such as anti-nuclear antibodies, as well as enabling a better appreciation of neural specific staining. In addition, the utilization of a pre-absorption step can reduce the amount of these non-specific antibodies. This improves the detection of lower titer antibodies and eliminates the unnecessary excessive reflex testing that may occur due to interfering non-specific antibodies. 

    The figure on the left is shown as a representation of some of the staining patterns of reportable neural antibodies, and demonstrates the diversity of staining patterns and some of the subtle differences between antibodies. The use of tissues from critical regions of the brain within a single composite is an essential aspect to this testing. The interpretation process itself requires extensive training and expertise, and should be augmented through the exposure of hundreds of well-defined clinical samples to refine the interpreter’s abilities. Although it's important to note that this is an extremely challenging task for most laboratories. Similar to biopsy tissue examination in a pathology laboratory where a critical diagnosis may rest on the pathologist’s interpretation, the improper identification of neural antibodies can have severe consequences, such as missing a treatable or even curable disease, where there is a limited opportunity for intervention. 

    Challenges for laboratories

    Given the critical importance of interpretation of staining patterns and neural antibody evaluations, it's extremely important that the testing laboratory build out a robust training program to maintain a high quality of interpretation. Unfortunately, this is particularly challenging for laboratories wishing to perform neural antibody testing, given several significant barriers. First, the prevalence of these disorders is low, making it difficult to acquire clinically defined positive samples to use as part of the training. Second, there's a lack of assay standardization, and no or limited access to national or international proficiency testing materials. To highlight the difficulties interpreting these assays, one of the few external quality assessment programs that focuses on neural antibodies found strong discordance across laboratories despite, in some cases, the use of the same kit or reagent — highlighting the challenges in interpretation of these staining patterns.

    So shown here in the figure at the bottom, for the NMDA receptor, which is one of the more critical antibodies that can be detected in these evaluations, there was disagreement across fourteen laboratories. When a positive confirmed case was shared with these fourteen laboratories, only five of the laboratories properly identified the NMDA staining pattern. The challenge here really lies in distinguishing non-specific or pattern mimics from the real staining. And this is particularly challenging for the NMDA receptor.

    While there are other factors at play unrelated to the interpretation, many laboratories do perform minimal validation verification of the performance of commercial kits with unlimited number of positive and negative control samples. This likely relates to the lack of availability of positive specimens and urgency to validate these assays quickly and, in some circumstances, an underestimation of the complexity of the interpretation and dependency on the claims by the manufacturer, where highly experienced individuals would have interpreted the assays, but this may not reflect the actual testing laboratory’s capabilities. 

    Lack of standardization

    The consequence of the lack of expertise and differences across laboratories and how testing for neural antibodies are approached can be substantial. Here is a case of a 61-year-old male who presented with a history of sensory neuronopathy, as evidenced by a gait and balance along with loss of vibrational sense. Initially, the patient was seen at their local hospital where a paraneoplastic evaluation was ordered at a commercial reference laboratory. The laboratory reported the presence of anti-Hu antibodies, also known as Anna-1 antibodies, at a relatively low titer. The methodology used was not provided to the ordering clinician. While the patient's presenting features were consistent with anti-Hu, or Anna-1, the long duration of symptoms in the patient raised questions about a diagnosis of Anna-1 paraneoplastic disorder. No small cell lung cancer was identified. The patient was subsequently evaluated at Mayo Clinic, where testing is repeated in our laboratory using both tissue immunofluorescence and a confirmatory western blot method, which were both negative. Subsequently, genetic testing was performed where our genetic diagnosis was reached unrelated to Anna-1 or Hu. 

    An important take-home message is that there is significant differences across laboratories in terms of the methodologies that are utilized, how those methods are organized into testing algorithms, and the quality of interpretation of tissue immunofluorescence and cell-based assays. This is further exacerbated by the lack of external quality assessment programs within the United States. The lack of assay standardization, and the lack of availability of reference material. 

    Given the mentioned limitations of neural antibody testing, in the Neuroimmunology Laboratory at Mayo Clinic not only do we evaluate standard analytical performance characteristics, such as precision accuracy, reportable range, reference range, and analytical specificity, we also incorporate a rigorous evaluation of diverse samples and ensure a high degree of interpretive competency is present in our staff before performing clinical testing. This is aided through the access to a large biorepository of specimens that can be utilized for these verification and validation studies. This is particularly important for assays that utilize subjective microscope based interpretation. 

    Importance of a robust training and competency program

    Given the critical importance of tissue immunofluorescence interpretation, and these neural antibody evaluations, it's extremely important that the testing laboratory build out a robust training program to maintain a high quality of interpretation. Our group at Mayo Clinic developed a nine week long training program to ensure that the staff interpreting tissue and cell-based immunofluorescence assays are able to correctly identify all reportable antibodies with high sensitivity and specificity. Furthermore, all positive cases are reviewed by a neurologist or a Ph.D.-level laboratory director prior to reporting out results.

    The training program itself utilizes a biobank of clinically defined specimens that are used to train staff. First in an unblinded fashion and then subsequently in a blinded fashion. The program includes written and microscope-based work. After this intensive nine-week program of training, there's a competency examination that's blinded. The individual will be asked to interpret a complex group of reportable neural antibodies, often in the presence of co-occurring antibodies. If the examination is passed the staff member then would be allowed to begin interpreting immunofluorescence and cell-based assays on a clinical basis.

    After this nine-week program, and the competency evaluation, we still maintain our competency through regular yearly evaluations of a reader's performance across all reportable antibodies. This is extremely important to ensure that the quality of the interpretation being provided to clients offers a high degree of confidence, as well as accuracy. 

    Benefits of an integrated approach

    A critical component of ensuring the quality of this testing is made possible through the laboratory utilizing an integrated approach between the Neuroimmunology Laboratory, our Mayo Clinic Autoimmune Neurology Clinic, as well as research and discovery labs. One benefit to this approach is the identification and clinical definition of samples that are used in these competency training programs, as well as being used as part of our test development process. This integrated approach also allows the laboratory to customize and provide clinically relevant interpretive comments that are vetted both by the laboratory, as well as clinicians. This approach also supports the discovery of novel antibodies. Importantly, this makes it possible to more readily review the performance of current clinical tests and determine quickly when they're not meeting the clinical need. Ultimately, this drives the laboratories ability to develop and implement methodological improvements. Particularly, as an example, I'm going to discuss aquaporin 4 (AQP4) and MOG antibody detection platforms. 

    AQP4 testing approaches

    So as an example, initially AQP4 antibodies, which are critical biomarker for neuromyelitis optica spectrum disorders (NMOSD), were discovered initially through unique staining pattern on tissue immunofluorescence. However, as the disorder was further characterized and new methods were developed and assessed, it was realized that this initial immunofluorescence approach had limited clinical sensitivity. And therefore additional assays were developed, evaluated, and offered clinically. After several iterations of these assays, where the performance was reviewed, determined that they weren't yet meaning clinical need, we finally arrived at the current gold standard approach to testing, which utilizes live cell assay formats, which we now know provide the highest diagnostic accuracy. AQP4 antibodies have thought to be pathogenic and recognize specific native AQP4 protein conformations, but which are best recapitulated in an assay by the use of live cells expressing the AQP4 protein on their cell surface. This allows us to offer the highest diagnostic accuracy. 

    MOG live cell-based assays (CBAs) outperform fixed or ELISA-based assays in large multicenter study

    This has led to a paradigm shift in the way we approach developing assays to detect antibodies targeting proteins expressed on the cell surface. Currently our laboratory utilizes live cell-based assays to also detect antibodies that bind to the myelin oligodendrocyte glycoprotein, known as MOG, which serve as a biomarker of MOG associated disorders, also known as MOGAD, which is a neuroinflammatory condition that clinically overlaps with NMOSD and multiple sclerosis. Recently, in a large international multi-center study, it was demonstrated that live cell assays provide higher diagnostic accuracy as compared to fixed cell-based assays, or those that use denatured MOG proteins, such as ELISAs.

    Shown on the left in this figure is the performance of a set of known positive samples and known negative samples that were tested by all CBAs, which included fixed CBAs as well as live cell CBAs. The detection rate is shown on the y-axis. In comparison, on the right side of this figure is the performance of only the live cell CBAs where you can see an increase in identification of the positive samples and the lack of identification of negative samples, indicating superior diagnostic accuracy. These studies indicated that approximately 10-15% of cases would be missed if relying solely on non-live cell assays or through the use of fixed assays. It is therefore recommended that if results obtained on these assays are conflicting with the clinical picture, testing on a live cell assay should be pursued.

    Currently, our laboratory is the only reference laboratory within the United States utilizing live cell-based assays to detect MOG antibodies. Another important take-home point from the study was there was relatively poor concordance across laboratories for weakly positive antibodies — further indicating a lack of standardization of assays and some of the more challenging aspects to subjective microscopy-based methods where there tends to be a higher degree of discordance across interpretations. 

    Cell-based assays considerations

    In summary, when considering the use of cell-based assays, there are important distinctions between these assay types. Generally, live cell-based assays are superior for the defection of cell surface antigen targets, as is exemplified by AQP4 and MOG, as compared to fixed cell-based assays. Microscopy-based assays that require reader interpretation have a higher degree of subjectivity and can lead to poor inter- or intra-laboratory agreement, meaning the same sample tested over and over, the interpretation may be different. This can be reduced or eliminated through the use of semi-quantitative flow cytometry based cell-based assays, such as used at the Mayo Clinic for the detection of AQP4 and MOG antibodies. 

    Conclusion

    In conclusion, tissue indirect immunofluorescence plays a central role in neural antibody testing. The expertise of those performing the interpretation of these assays is critical. Given the complexity of testing, a comprehensive assessment of assay performance in-house should be performed by the testing laboratory prior to or offering these assays clinically. Laboratories should not depend on performance claims by the manufacturer and only utilize a limited number of samples to ensure that they're performing the assays properly. Access to clinically defined samples is essential both for a  laboratory’s ability to train staff, but also to complete an appropriate and thorough test verification and validation study. It's always important to enquire about who is interpreting the results. The degree and the quality of the interpretation is critical to the quality of the assay.

    References

    1. Budhram A, Dubey D, Sechi E, et al. Neural antibody testing in patients with suspected autoimmune encephalistis. Clin Chem. 2020 Nov 22. Epub ahead of print.
    2. Waters PJ, Pittock SJ, Bennett JL, Jarius S, Weinshenker BG, Wingerchuk DM. Evaluation of aquaporin-4 antibody assays. Clin Exp Neuroimmunol.2014;5(3):290-303.

    MCL Education

    MCL Education

    This post was developed by our Education and Technical Publications Team.