IgG production within the central nervous system: a critical review of proposed formulae.

Demonstration of intrathecal IgG production is employed in the diagnosis of various neurological disorders. This pathological IgG fraction in cerebrospinal fluid (CSF) can be visualized directly as oligoclonal bands by electrophoresis or isoelectric focusing or can be calculated as "excess" or "synthesized" IgG according to different formulae. A comparison of the results obtained with isoelectric focusing and with five formulae showed that even though three of the formulae discriminated well between a reference population and patients with multiple sclerosis, all five gave wrong and misleading results in the presence of blood-brain barrier damage, as defined by an abnormally raised CSF/serum albumin ratio. A mathematical and statistical evaluation of the different formulae showed only those based on covariance between CSF/serum IgG and CSF/serum albumin to be valid, and these only when values of CSF/serum albumin were normal. Among the five formulae the IgG index (equal to CSF/serum IgG:CSF/serum albumin) is unique in having a comparatively small and constant maximal relative error resulting from the variation coefficients of the IgG and albumin assays. In the case of blood-brain barrier damage, there exists currently no valid procedure to calculate intrathecally produced IgG; in such instances sensitive electrophoretic or isoelectric focusing methods demonstrating oligoclonal IgG bands are most appropriate to demonstrate intrathecal IgG production.