Thiomersal : neurobehavioural studies in animal models
The Committee considered whether animal models could be applied in order to better understand the association, if any, between thiomersal (containing ethyl mercury) and neurobehavioural disorders in infants, children and adults receiving thiomersal-containing vaccines. It was noted that in the neuropathology of autism - one neurobehavioural disorder that has received much public attention - the characteristic features of increased brain weight, total brain volume, cortical grey matter volume and neuronal cell density in the limbic system, with a decreased number of Purkinje cells in the cerebellum and absence of gliosis were not consistent with an external toxic agent. There is no evidence in autism of neurodegeneration, a pathological feature that would be expected if the cause of the disease were toxic.
From an expert presentation made to the Committee and from several publications, it is clear that: (i) no precise animal model exists that closely mimics autism in humans, although animal models of deficit in social play do exist; (ii) in the models available, susceptibility to neurobehavioural disorders has a genetic basis; (iii) there are experimental data to suggest that there is a link between autoimmune deficiency and predisposition to autism (although this remains conjectural); and (iv) mice born to mothers infected with human influenza virus have developed neuropathologies similar to those described in association with autism.
The Committee further reviewed results published in the scientific literature analysing a potential neuropathological effect of ethyl mercury in various strains of mice. Although a particular autoimmune disease-sensitive strain of mice showed some neurological and behavioural changes following administration of thiomersal, it was concluded by the Committee that the general picture presented did not mimic autism in humans.
The Committee identified two ways to further investigate the safety of thiomersal: epidemiological studies on the effects of ethyl mercury and pharmacokinetic studies in infants. Little is known of any particular susceptibility to thiomersal toxicity in infants weighing less than 2.5 kg, and in malnourished infants. Epidemiological studies are in progress. Special consideration is being given to how predisposition to thiomersal toxicity in infants weighing less than 2.5 kg and in infants with malnutrition might be addressed.