Will anti-neuronal antibody biomarkers make neuropsychiatric systemic lupus erythematosus no longer a diagnosis of exclusion?

Abstract

Development of the American College of Rheumatology (ACR) criteria for neuropsychiatric systemic lupus erythematosus (NPSLE) was a major step in understanding these disorders. Without biomarkers confirming a direct relationship between systemic lupus erythematous and the neuropsychiatric diagnoses, skeptics may view the ACR classification as mere semantics. However, by allowing neurologists and rheumatologists in different centers to use standardized terms, the criteria are useful in caring for patients and in assembling homogeneous groups of research participants. In pediatric systemic lupus erythematous cohorts, the prevalence of NPSLE in general, and movement disorders in particular, ranged from 34.6% and 0% in a retrospective study in Taiwan1 to 95% and 7% in a prospective study in the United States.2 For many patients, diagnosis and treatment are delayed as mimicking infectious, toxic, and metabolic conditions are ruled out. A diagnostic biomarker would be immensely helpful. Although history, examination, magnetic resonance imaging, cerebrospinal fluid (CSF) analysis, and standard autoantibody profiles are important, none are sufficiently sensitive or specific to be considered a diagnostic biomarker for NPSLE. The search for informative biomarkers in neuro-immunological conditions including NPSLE has been hampered by the use of techniques which alter the target antigen’s structure. Studies using cell-based assays, which preserve antigenic structure, have produced more clinically useful and scientifically valid results. At least two prior studies have used cell-based, flow cytometry assays in NPSLE. In adult patients with NPSLE, Alosachie et al.3 found antibodies to a neuroblastoma cell line in 63% (36/57) of serum samples and 83% (5/6) of CSF samples, but none using a glioma cell line. In a cohort of 17 adult patients with NPSLE, Kang et al.4 detected antibodies to the same neuroblastoma cell line in 77% (13) of serum and 88% (15) of CSF samples. In the latter study, abnormalities in magnetic resonance imaging and CSF anti-ribosomal P antibody assays were found in only 59% and 18% of patients with NPSLE respectively. In both studies, patients with systemic lupus erythematous without neuropsychiatric symptoms, and disease and healthy controls rarely harbored such antibodies, thus yielding good positive and negative predictive values, particularly for the CSF assays.3,4 Dale et al.,5 the first group to apply these techniques to pediatric NPSLE patients, found elevated serum immunoglobulin G binding to the surface of cells with dopaminergic neuronal characteristics in six females with chorea (n=4) or parkinsonism (n=2) associated with lupus and/or anti-phospholipid antibodies compared with healthy and disease controls. The fact that similar levels of antibodies were detected in patients with lupus without neuropsychiatric symptoms contrasts with the prior reports and is somewhat troubling. This finding suggests a difference between pediatric and adult NPSLE and/or the requirement for a ‘second hit’ to cause clinical expression, but ultimately remains unexplained in the current study. From the above studies, several points emerge. First, anti-neuronal autoantibodies appear to be important in NPSLE. Secondly, based on the two adult studies, measurement of such antibodies in CSF may be more meaningful than in serum. The lack of testing in CSF in the current study is an important limitation. Lastly, the target cell type chosen for the antibody assays may be critically important. An elegant feature of the current study is that the authors used a cell line with dopaminergic characteristics to detect antibodies in the subgroup of patients with movement disorders within NPSLE. This raises several questions. Does this approach explain why Dale et al. found elevated antibodies in all of the patients, compared with prior studies with lower sensitivity? Would the dopaminergic cell assay detect antibodies in other NPSLE subgroups? What is the best cell type to use for other, more common NPSLE diagnoses, such as headaches, seizures, and mood disorders? Future research will hopefully identify the specific antigen targeted by the autoantibodies in patients with NPSLE, perhaps within diagnostic subgroups. Such identification is necessary to elevate the status of the currently identified autoantibodies from an informative, but not definitive, finding, to the next big diagnostic biomarker in the field.