Background: No consensual criteria exist to diagnose neuroborreliosis. The intrathecal anti-Borrelia antibody index (AI) is a necessary criterion to diagnose neuroborreliosis in Europe, but not in the United States. Previous studies to determine the diagnostic value of the AI found a sensitivity ranging from 55% to 80%. However, these studies included only typical clinical cases of meningitis or meningoradiculitis, and none had a control group with CSF anti-Borrelia antibodies.
Methods: We studied a sample of 123 consecutive patients with clinical signs of neurologic involvement and CSF anti-Borrelia antibodies. We determined the AI for all patients and a final diagnosis was made. Patients were then divided into three groups (neuroborreliosis, possible neuroborreliosis, control).
Results: Thirty of the 40 patients with neuroborreliosis had a positive AI (AI sensitivity = 75%). Two of the 74 patients with another neurologic diagnosis had a positive AI (AI specificity = 97%).
Conclusion: The antibody index has a very good specificity but only moderate sensitivity. Given the lack of consensual criteria for neuroborreliosis and the absence of a "gold standard" diagnostic test, we propose pragmatic diagnostic criteria for neuroborreliosis, namely the presence of four of the following five items: no past history of neuroborreliosis, positive CSF ELISA serology, positive anti-Borrelia antibody index, favorable outcome after specific antibiotic treatment, and no differential diagnosis. These new criteria will need to be tested in a larger, prospective cohort.