Kryzer, Thomas J.

Abstract Background Ma-2 is a neuronal nucleolar protein with purported involvement in RNA transcription and apoptosis regulation. Patients with Ma2 antibodies present with autoimmune encephalitis (limbic or brainstem). Ma-2 antibodies are a high-risk marker for paraneoplastic neurologic cause for autoimmune encephalitis (>75% of patients having either testicular cancer or non-small cell lung cancer). In older patients, there is a higher frequency of small cell lung cancer or other cancers especially when Ma 1 antibodies are also detected. Identification of Ma2 antibodies is important to help diagnose autoimmune encephalitis, guide the cancer search and aid treatment selection. The objective of this study was to develop and document the performance of a qualitative enzymatically amplified sandwich-type Ma2 Antibody ELISA. Methods The Ma2 ELISA was produced in-house utilizing commercially available protein and was performed on a Hamilton Star liquid handling instrument equipped with a plate washer and plate reader. Each assay plate included patient samples along with 3 independent negative serum calibrators and 4 quality controls (Both positive and negative controls per matrix [CSF or serum]) each tested in duplicate. Mean optical densities (OD) for patient specimens were compared to the average OD of the negative serum calibrators to generate a ratio. A ratio cut-off of 8 was selected for reporting positivity based on preliminary studies and was validated during the method performance assessment. To assess assay performance we measured precision (3 positive pools at or above the cut-off ratio for each specimen type were tested 20 times in one assay and in duplicate across 10 independent assays), accuracy comparing to the current reference method at Athena® Diagnostics (N = 10 positives / 40 negatives), analytical specificity (20 samples with high titer autoantibodies directed against other proteins), and the impact of common interferences (lipids, bilirubin, hemolysate were spiked into 2 positive and 1 negative sample per interference). Clinical performance of the assay for differentiating healthy donors without disease (N = 140) and those with non-Ma2 neurological autoimmunity (N = 23) from those with clinical phenotypes consistent with Ma-2 associated encephalitis (N = 8) was measured. Results Serum intra-assay imprecision (coefficient of variation; %CV) ranged from 3.0–6.5%. Inter-assay imprecision ranged from 21–29%. CSF intra-assay imprecision ranged from 3.5–10%. Inter-assay imprecision ranged from 13–15.5%. Accuracy was assessed via qualitative comparison to a reference method [western blot or nanoliter scale immunoassay]. There was perfect agreement between methods. To assess analytical specificity, samples with high concentrations of immunoglobulins (hypergammaglobulinemia [N = 10] or Systemic Lupus Erythematosus [N = 10]) were tested in serum (and spiked into CSF); all were negative. Interference studies were conducted on 2 positive Ma2 serum and CSF samples, spiked for highly hemolytic (1000 mg/dL), lipemic (2000 mg/dL) and icteric (60 mg/dL) conditions. All samples remained positive under given interferents. A negative sample was tested for each matrix, both remained negative. Clinical specificity was 100% with all samples from patients without Ma-2 encephalitis testing negative (N = 183). All ELISA positives identified in this study had phenotypes consistent with Ma-2 autoimmune encephalitis (N = 8). Conclusion The performance of the Ma2 ELISA was acceptable for clinical laboratory implementation.

Objective This study was undertaken to describe a novel biomarker of germ cell tumor and associated paraneoplastic neurological syndrome (PNS). Methods Archival sera from patients with germ cell tumor–associated PNS were evaluated. We identified a common autoantigen in a human testicular cancer cell line (TCam‐2) by Western blot and mass spectrometry. Its identity was confirmed by recombinant‐protein Western blot, enzyme‐linked immunosorbent assay (ELISA), and cell‐based assay. Autoantibody specificity was confirmed by analyzing assorted control sera/cerebrospinal fluid. Results Leucine zipper 4 (LUZP4)–immunoglobulin G (IgG) was detected in 28 patients' sera, 26 of whom (93%) were men. The median age at neurological symptom onset was 45 years (range = 28–84). Median titer (ELISA) was 1:300 (1:50 to >1:6,400, normal value < 1:50). Coexistent kelchlike protein 11–IgG was identified in 18 cases (64%). The most common presenting phenotype was rhombencephalitis (17/28, 61%). Other presentations included limbic encephalitis (n = 5, 18%), seizures and/or encephalitis (n = 2, 7%), and motor neuronopathy/polyradiculopathy (n = 4, 14%). The most common malignancy among cancer‐evaluated PNS patients was seminoma (21/27, 78%). Nine of the 21 seminomas detected by whole‐body fluorodeoxyglucose positron emission tomography scan (43%) were extratesticular. Both female patients had ovarian teratoma. Regressed testicular germ cell tumors were found in 4 patients. Exposure of T‐cell–dendritic‐cell cocultures from chronic immunosuppression‐naïve LUZP4‐IgG–seropositive patients to recombinant LUZP4 protein evoked a marked increase in CD69 expression on both CD4+ and CD8+ T cells when compared to vehicle‐exposed and healthy control cultures. Interpretation LUZP4‐IgG represents a novel serological biomarker of PNS and has high predictive value for germ cell tumors. The demonstrated antigen‐specific T‐cell responses support a CD8+ T‐cell–mediated cytotoxic paraneoplastic and antitumor potential. ANN NEUROL 2021;89:1001–1010