Neuronal autoantibodies: differentiating clinically relevant and clinically irrelevant results

The aim of this study is to compare the rates of clinically relevant and clinically irrelevant neuronal autoantibodies among patients presenting with new neurological symptoms. We reviewed 401 neurological patients who were tested for the Mayo-Clinic paraneoplastic panel from January 2014 to December 2014 at the Johns Hopkins Hospital. We divided antibody-positive patients into two groups: clinically relevant (CR), in which a recognizable autoimmune or paraneoplastic syndrome was confirmed, and clinically irrelevant (CI), in which an autoimmune/paraneoplastic etiology was initially suspected but an alternative diagnosis was eventually found. We used Fisher's exact test for categorical variables and Mann-Whitney U test for continuous variables to identify differences between the two groups. Fifty-three patients tested positive for one or more neuronal autoantibodies. There were 17 CR (65% females, mean age 56 years), 33 CI, and 3 indeterminate patients. Compared to CI patients, CR patients were more likely to present with movement disorders or stiff person syndrome, have inflammatory CSF markers, cancer or smoking history, concomitant hyponatremia, and classical onconeuronal antibodies. CI patients were more likely to have a neuromuscular presentation, a chronic course, and antibodies against synaptic antigens. By combining the most robust differentiating factors, we developed a simple scale that predicted clinical relevance with an odds ratio of 50.3 (CI 8.2-309.9, P < 0.0001) if the score was >= 2. Up to 62% of neuronal autoantibody-positive patients are ultimately found to have an alternative diagnosis. Several clinical and laboratory factors can differentiate CR from CI patients to aid in interpretation of positive results.