Clinical Sensitivity, Specificity, and Predictive Value of Neural Antibody Testing for Autoimmune Encephalitis

Abstract

A plethora of antibodies against neural antigens have emerged as biomarkers of autoimmune encephalitis. This has led to a rise in neural antibody testing among patients suspected of having this disease, which has in turn motivated evaluations of its diagnostic utility in clinical practice. We discuss clinical sensitivity, specificity, and predictive value of neural antibody testing for autoimmune encephalitis, with the aim of promoting appropriate test implementation and interpretation. Clinical sensitivity refers to a test’s ability to correctly identify individuals with a disease; the higher a test’s clinical sensitivity, the fewer false negatives it generates. Meanwhile, clinical specificity refers to a test’s ability to correctly exclude individuals who do not have the disease; the higher a test’s clinical specificity, the fewer false positives it generates. These diagnostic test measures are intrinsic to the assay and independent of the disease prevalence in the tested population (in contrast to predictive value, discussed later). Clinical sensitivity is challenging to determine for neural antibody testing because independent diagnostic gold standards to identify false-negative results in patients with the disease of interest are typically lacking. Thus, attempts to calculate clinical sensitivity for a given neural antibody often use the clinical syndrome to define what constitutes a false-negative result. For example, cell-based assays (CBAs) for detection of myelin oligodendrocyte glycoprotein (MOG)-IgG, which is associated with acute disseminated encephalomyelitis, optic neuritis, and myelitis, had an initial reported clinical sensitivity of only 25%; this was calculated by assuming a MOG-IgG–compatible clinical phenotype to be the diagnostic gold standard (1). Instinctively, however, not every MOG-IgG–negative patient whose clinical presentation falls within the spectrum of MOG-IgG–associated disease (e.g., acute disseminated encephalomyelitis, optic neuritis, or myelitis) is a false negative, and numerous other neurological diseases may present similarly (e.g., multiple sclerosis, neurosarcoidosis, neurosyphilis, primary central nervous system lymphoma). Given this inherent limitation to determining absolute clinical sensitivity of a given neural antibody, greater emphasis has been placed on relative clinical sensitivity across assays or fluid specimens (see Table 1). Test methodology and practical considerations for neural antibody testing in autoimmune encephalitis. See online Data Supplement for table references. In general, paired serum and CSF submission is recommended in patients with suspected autoimmune encephalitis to maximize clinical sensitivity and specificity. VGKC antibody testing is being phased out due to low clinical sensitivity/specificity of VGKC antibodies alone for anti-LGI1/CASPR2-positive neurological autoimmunity. Abbreviations: AMPAR, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor; ANNA-1, antineuronal nuclear antibody type 1; ANNA-2, antineuronal nuclear antibody type 2; AQP4, aquaporin-4; CASPR2, contactin-associated protein-like 2; CBA, cell-based assay; CRMP5, collapsin response mediator protein-5; CSF, cerebrospinal fluid; DNER, delta/notch-like epidermal growth factor–related receptor; DPPX, dipeptidyl-peptidase–like protein 6; FACS, fluorescence-activated cell sorting; GABAAR, gamma-aminobutyric acid (A) receptor; GABABR, gamma-aminobutyric acid (B) receptor; GAD65, glutamic acid decarboxylase-65; GFAPα, glial fibrillary acidic protein alpha isoform; GlyRα1, glycine receptor alpha-1 subunit; IB, immunoblot; IF, indirect immunofluorescence; IgLON5, immunoglobulin-like cell adhesion molecule 5; KLHL11, Kelch-like protein 11; LGI1, leucine-rich glioma-inactivated 1; MAP1B, microtubule-associated protein 1B; mGluR1, metabotropic glutamate receptor 1; mGluR5, metabotropic glutamate receptor 5; MOG, myelin oligodendrocyte glycoprotein; NMDAR, N-methyl-D-aspartate receptor; PCA-1, Purkinje cell cytoplasmic antibody type 1; PCA-2, Purkinje cell cytoplasmic antibody type 2; PCA-Tr, Purkinje cell cytoplasmic antibody type Tr; T1DM, type 1 diabetes mellitus; TIIF, tissue indirect immunofluorescence; VGKC, voltage-gated potassium channel; WB, western blot. Test methodology and practical considerations for neural antibody testing in autoimmune encephalitis. See online Data Supplement for table references. In general, paired serum and CSF submission is recommended in patients with suspected autoimmune encephalitis to maximize clinical sensitivity and specificity. VGKC antibody testing is being phased out due to low clinical sensitivity/specificity of VGKC antibodies alone for anti-LGI1/CASPR2-positive neurological autoimmunity. Abbreviations: AMPAR, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor; ANNA-1, antineuronal nuclear antibody type 1; ANNA-2, antineuronal nuclear antibody type 2; AQP4, aquaporin-4; CASPR2, contactin-associated protein-like 2; CBA, cell-based assay; CRMP5, collapsin response mediator protein-5; CSF, cerebrospinal fluid; DNER, delta/notch-like epidermal growth factor–related receptor; DPPX, dipeptidyl-peptidase–like protein 6; FACS, fluorescence-activated cell sorting; GABAAR, gamma-aminobutyric acid (A) receptor; GABABR, gamma-aminobutyric acid (B) receptor; GAD65, glutamic acid decarboxylase-65; GFAPα, glial fibrillary acidic protein alpha isoform; GlyRα1, glycine receptor alpha-1 subunit; IB, immunoblot; IF, indirect immunofluorescence; IgLON5, immunoglobulin-like cell adhesion molecule 5; KLHL11, Kelch-like protein 11; LGI1, leucine-rich glioma-inactivated 1; MAP1B, microtubule-associated protein 1B; mGluR1, metabotropic glutamate receptor 1; mGluR5, metabotropic glutamate receptor 5; MOG, myelin oligodendrocyte glycoprotein; NMDAR, N-methyl-D-aspartate receptor; PCA-1, Purkinje cell cytoplasmic antibody type 1; PCA-2, Purkinje cell cytoplasmic antibody type 2; PCA-Tr, Purkinje cell cytoplasmic antibody type Tr; T1DM, type 1 diabetes mellitus; TIIF, tissue indirect immunofluorescence; VGKC, voltage-gated potassium channel; WB, western blot. When considering the clinical sensitivity of neural antibody testing for autoimmune encephalitis, it is important to remember that phenotypic overlap between antibody-associated encephalitides can be substantial (2). Implementing panels that test for multiple neural antibodies is therefore recommended to maximize clinical sensitivity and reduce the likelihood of missing a clinically actionable result (3). However, calculating clinical sensitivity of neural antibody panel testing for autoimmune encephalitis is hindered by the fact that not all autoimmune encephalitides are antibody-associated; it thus follows that a negative neural antibody panel in a patient with autoimmune encephalitis is not necessarily a false-negative result (4). For this reason, thorough clinical evaluation is required to rule out autoimmune encephalitis in an antibody-negative patient. Algorithms designed to aid in the diagnosis of autoimmune encephalitis, independent of neural antibody status, are helpful in this regard (4). Clinical specificity of neural antibody testing for autoimmune encephalitis is even more of a concern to the laboratorian than clinical sensitivity. This is because a false-positive neural antibody result may have severe consequences to patient care; it can lead to potentially hazardous immunotherapy administration, repeated unnecessary malignancy screenings, and missed treatable alternative diagnoses (4–6). Clinical specificity for some neural antibodies can be maximized by first using brain tissue indirect immunofluorescence (TIIF) as a screening assay, with performance of an appropriate second assay (e.g., western blot/immunoblot or CBA) to confirm antibody positivity if characteristic staining on TIIF is observed (see Table 1) (3, 7). This 2-step approach is beneficial to clinical specificity but can be challenging to implement in laboratories that lack the expertise to interpret TIIF or the resources to institute more sophisticated reflex-based testing algorithms. This has prompted some laboratories to use commercialized, traditionally confirmatory assays (e.g., immunoblot, CBA) in isolation as a way of screening for neural antibodies (5). While such commercial kits can improve test accessibility and turnaround times for patients when used in isolation, laboratorians should be aware that this approach lowers clinical specificity and increases the possibility of false-positive results (7–9). In cases of positivity by a single commercial assay in isolation, this information should be relayed to the ordering clinician to encourage clinical-serological correlation and minimize risk of misdiagnosis due to a false-positive result. In some patients, autoimmune encephalitis is a remote immune-mediated effect of an underlying tumor and is termed “paraneoplastic” (10). Neural antibodies may be detected in such cases that have clinical specificity for both paraneoplastic autoimmune encephalitis and an underlying tumor; the detection of either should therefore be considered a true-positive result. This takes on importance in the rare instance that a patient has neural antibody testing performed for suspected autoimmune encephalitis and an antibody is detected that is strongly linked to cancer (i.e., >70% cancer association, termed “high-risk” antibody) (10), but a more likely alternative diagnosis for their neurological symptoms is determined (e.g., metabolic derangement, neurodegenerative disease). While the neural antibody may represent a false-positive result in the absence of autoimmune encephalitis, the possibility that it is a true-positive result indicating an underlying occult tumor remains (11). For this reason, comprehensive cancer screening should be contemplated in a patient with a high-risk antibody regardless of the clinical presentation, especially after assessment of patient factors such as smoking status and malignancy history. Some antibodies may have a more variable association with both neurological autoimmunity and malignancy when interpreted in isolation, such as low-titer P/Q-type and N-type voltage-gated calcium channel antibodies, highlighting the need for ongoing evaluations of their inclusion in autoimmune encephalitis panels to maximize clinical specificity (12). Through discussion of clinical sensitivity and specificity, it becomes clear that recalling details pertaining to these diagnostic measures for each neural antibody can be daunting. Interpretative comments alongside antibody results, including any concerns regarding clinical sensitivity or specificity depending on the test methodology (e.g., single assay, 2-step TIIF-based reflex testing with second confirmatory assay, assay with diagnostic cutoff that impacts likelihood of neurological autoimmunity) or type of specimen received [i.e., serum or cerebrospinal fluid (CSF)] are thus immensely helpful to the clinician and should be standardly included in laboratory reports. Negative predictive value (NPV) refers to the probability a patient truly does not have the disease if a test is negative, while positive predictive value (PPV) refers to the probability a patient truly does have the disease if a test is positive. Unlike clinical sensitivity and specificity, which are intrinsic to the assay and independent of the tested population, NPV and PPV of a test are impacted by disease prevalence. The NPV of a test increases as the prevalence of the disease in the tested population decreases. The prevalence of autoimmune encephalitis is low (13), which benefits the NPV of neural antibody testing because the majority of tested patients are thus true negatives. In contrast, the PPV of a test decreases as the prevalence of the disease in the tested population decreases. As availability of neural antibody testing has increased, so too has its ordering among patients with a wide range of neurological symptoms that may not be typical of autoimmune encephalitis. Neural antibody testing in patient populations with a low likelihood of neurological autoimmunity is liable to generate false positives, a problem that is exacerbated by suboptimal test clinical specificity. As an example, detection of certain neural antibodies (e.g., anti-Yo) using commercial immunoblot kits in isolation, rather than as confirmatory assays after screening by TIIF, is a testing approach that is utilized in some laboratories despite its lower clinical specificity. While this reduction in clinical specificity may be marginal, it can lead to a dramatic decline in PPV when the test is applied to a population with low disease prevalence. The clinical specificity of neural antibody testing using commercial immunoblot kits alone has been estimated to still be relatively high (near 95%), yet the PPV of this approach in studies is only 30% to 50%; this is at least in part due to broadening use of neural antibody testing among patients who are unlikely to have autoimmune neurological disease (5–8). This potential for false-positive or clinically irrelevant results when testing in low-probability scenarios has also been reported for neural antibodies detected by various other methodologies, including serum anti-N-methyl-D-aspartate receptor, anti-contactin-associated protein-like 2, anti-MOG, and high-titer antiglutamic acid decarboxylase-65 (9, 14, 15). Appropriate patient selection when ordering neural antibody testing is therefore essential to minimize the risk of false-positive results, and a positive neural antibody result in a patient with an atypical presentation for neurological autoimmunity (e.g., insidious disease progression, normal ancillary testing including neuroimaging, electroencephalography, and CSF and/or presence of a more likely alternative diagnosis) should be scrutinized (4). In such cases, communication between the testing laboratory and ordering clinician to discuss the utility of additional evaluations (e.g., determination of serum positivity by a second assay, corroboration of serum positivity with CSF, retesting serum at a higher dilution to determine if positivity persists) is recommended. Neural antibody testing plays a major role in the diagnostic evaluation of patients with suspected autoimmune encephalitis. Through reviewing the diagnostic measures of clinical sensitivity, specificity, and predictive value, it becomes apparent that a neural antibody test result cannot be considered in isolation. An understanding of how the disease state is defined in relation to the neural antibody, knowledge of the test methodology implemented, and estimation of the pretest probability that the patient’s presentation is autoimmune are all required in each case to ensure appropriate test interpretation. Supplemental material is available at The Journal of Applied Laboratory Medicine online. Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved. Authors’ Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest: Employment or Leadership: None declared. Consultant or Advisory Role: None declared. Stock Ownership: None declared. Honoraria: None declared. Research Funding: D. Dubey, grants from Center of Multiple Sclerosis and Autoimmune Neurology. Expert Testimony: None declared. Patents: J.R. Mills, patent 10,753,945 with royalties paid to The Binding Site; D. Dubey, patent Kelch-like protein 11 pending, and a patent Leucine zipper 4 pending. Other Remuneration: D. Dubey, other from Grifols Pharmaceuticals, other from UCB Pharmaceuticals, outside the submitted work, other from Center of Multiple Sclerosis and Autoimmune Neurology.

Topics & Concepts