Background In the elderly, a decline in cognitive function is often associated with dementia. However, subacute deterioration on memory functions, associated with affective change and/or seizures are also classical features of limbic encephalitis, which is caused by inflammation of the limbic areas of the brain. Paraneoplastic limbic encephalitis is caused by the production of autoantibodies against neuronal tissues secondary to certain cancers. Similar autoantibodies have also been noted in the absence of neoplasms in non-paraneoplastic limbic encephalitis (nPLE), commonly reported antibodies being produced against Voltage Gated Potassium Channels (VGKC), which results in neuropsychiatric symptoms. This study aims to systematically review the evidence of the effects of treating VGKC associated limbic encephalitis in the elderly with intravenous Immunoglobulins, plasma exchange or steroids.

Method A systematic review of the available published and unpublished data were conducted, using the MEDLINE, EMBASE and PsychINFO databases, using pre-defined inclusion and exclusion criteria. Information such as patient demographics, presenting symptoms, treatment and antibody levels were gathered using a standardised data collection tool. The results were statistically analysed for correlation between clinical outcome and antibody titres before and after treatment.

Results A total of 59 cases met the inclusion criteria of whom 50 cases had complete data for statistical analysis. Twenty-one cases were over the age of 65 with complete data sets. Patients who clinically improved had the largest decrease in antibody titres. Patients presenting with affective and consistent neuroradiological changes were more likely to have an improved outcome following treatment. These results proved to be statistically significant. Statistical analysis also showed that 67.6% of patients below the age of 65 had an MRI change compared to 80% of patients above the age of 65. However, the observation was not statistically significant (p=0.292 χ² test).

Conclusion The overall evidence suggests that treatment of VGKC associated nPLE is effective. There are currently no agreed diagnostic guidelines, validated assessment tools or randomised controlled trials of the different treatment options. nPLE is a treatable condition which should form part of the differential diagnosis when patients present with subacute cognitive impairment, associated with affective changes and seizures.