Guillain-Barré syndrome and influenza virus infection.

Abstract

In Western countries, the cause of 60% of all Guillain-Barré syndrome (GBS) cases remains unidentified. The number of cases of unidentified cause peaks in winter, and these cases are commonly preceded by respiratory tract infection or influenza-like illness. We investigated the triggering role of influenza virus infection. Of 405 patients with GBS who were admitted to a French reference center during 1996-2004, 234 had cases caused by an unidentified agent. We used time-series methods to study the correlation between the monthly incidence of such cases and influenza-like illnesses reported by the Sentinelles surveillance network. We analyzed anti-influenza antibodies using complement fixation testing and hemagglutination-inhibition assays. We studied etiological subgroups using Wilcoxon and Fisher's exact tests. We found a positive association between the monthly incidence of GBS caused by an unidentified agent and reported influenza-like illnesses. Of 73 patients whose cases occurred during periods in which there was a possible link to influenza, 10 (13.7%) had serological evidence of recent influenza A, and 4 (5.5%) had serological evidence of influenza B. Eight of 10 influenza A-related cases occurred during "major" influenza seasons, and antibodies specific to the current epidemic strain were found in 9 cases. Most patients with influenza A-related cases were aged < 65 years, and none had antiganglioside antibodies. Influenza-related cases differed both from Campylobacter jejuni-related cases, with regard to the lack of need for mechanical ventilation (P = .014), and from the cases caused by an unidentified agent, with regard to the presence of preceding influenza-like illness or respiratory tract infection (P = .015) and longer time from the infectious event to GBS onset (P = .04). Influenza viruses are infrequent triggering agents of GBS but may play a significant role during major influenza outbreaks. Influenza-related GBS displays specific features and is not associated with antiganglioside antibody response, which suggests the presence of underlying immune mechanisms.