Thimerisol Study

\\

CONGRESSIONAL TESTIMONY

U.S. HOUSE OF REPRESENTATIVES

THE SUBCOMMITTEE ON HUMAN RIGHTS AND WELLNESS

COMMITTEE ON GOVERNMENT REFORM

SEPTEMBER 8, 2004 HEARING

TRUTH REVEALED:

NEW SCIENTIFIC DISCOVERIES

REGARDING MERCURY

IN MEDICINE AND AUTISM

submitted by:

Mady Hornig, MD

Director of Translational Research

Jerome L. and Dawn Greene Infectious Disease Laboratory

and

Associate Professor of Epidemiology

Mailman School of Public Health

Columbia University

Chairman Burton, Congressman Watson, and Members of the Subcommittee,

Thank you for the opportunity to submit for the record this statement regarding

our new animal model of the toxicity of thimerosal (ethylmercury preservative in

vaccines) and its implications for human health. I regret that I am unable to

personally present this testimony today due to a family medical emergency. Our

work addresses whether genes are important in determining if mercury

exposures akin to those in childhood immunizations can disrupt brain

development and function. I also submit for the record an electronic copy of the

first paper published on this animal model in the Nature Publishing Group journal,

Molecular Psychiatry (Hornig M, Chian D, Lipkin WI. Neurotoxic effects of

postnatal thimerosal are mouse strain dependent. Mol Psychiatry 2004;9:833-

845).

The premise of our research is that if mercury in vaccines creates risk for

neurodevelopmental disorders such as autism, genetic differences are likely to

contribute to that risk. We built upon an extensive, existing literature on toxicity of

other forms of mercury in inbred mouse strains that affirmed the importance of

specific genes controlling immune responses (major histocompatibility complex,

or MHC) in determining mercury-induced autoimmune outcomes in mice. Earlier

studies, however, did not use the form of mercury present in vaccines, known as

thimerosal, and did not consider whether intramuscular, repetitive administration

Page 2

during early postnatal development, when the brain and immune systems are still

maturing, might intensify toxicity. Based on reports of immune disturbances and

family history of autoimmune disease in a subset of children with autism, we

hypothesized that immune response genes linked to mercury immunotoxicity in

mice would predict damage following low-dose, vaccine-based mercury in our

mouse model.

Our predictions were confirmed. Using thimerosal dosages and timing that

approximated the childhood immunization schedule, our model of postnatal

thimerosal neurotoxicity demonstrated that the genes in mice that predict

mercury-related immunotoxicity also predicted neurodevelopmental damage.

Features reminiscent of those observed in autism occurred in the mice of the

genetically sensitive strain, including: generalized behavioral impoverishment

and abnormal reaction to novel environments; enlargement of the hippocampus,

a region of the brain involved in learning and memory; correlation of hippocampal

enlargement with abnormalities in exploration and anxiety; increased packing

density of neurons in hippocampus; and disturbances in glutamate receptors and

transporters. Only mice carrying the H-2

susceptibility gene showed these

autism-like effects (SJL/J mice). Two mouse strains with different H-2 genes

(C57BL6/J mice, H-2

; BALB/cJ mice, H-2

) did not demonstrate adverse

consequences following thimerosal exposure.

It is important to emphasize that these animal model studies do not provide

conclusive evidence regarding a link between mercury exposure and human

autism. Nonetheless, the finding that a specific genetic constraint profoundly

alters the brains and behavior of thimerosal-exposed mice confirms the biological

plausibility of thimerosal neurotoxicity, provides critical guidance for the

interpretation of existing epidemiologic investigations into the potential

association of thimerosal with neurodevelopmental disorders, and suggests

important new avenues for future research. Our work implies that if genetic

factors are operative in mediating a link between thimerosal and autism in

humans, then studies that fail to consider genetic susceptibility factors will be

compromised in their ability to detect a statistically significant effect even if one

exists.

Recent findings, presented at scientific meetings but as yet unpublished, suggest

that thimerosal neurotoxicity in susceptible mice involves the generation of

autoantibodies targeting brain components. This autoimmune response persists

long after the presence of mercury can no longer be detected. If confirmed, these

findings will enable us to develop a human diagnostic test to determine whether

some individuals with autism have similar autoantibodies present in their

peripheral blood. Such work would not only bring us a step closer to identifying

the genes associated with thimerosal neurotoxicity in humans, facilitating

prevention programs, it would also validate the utility of this animal model for the

development of safe and effective modes of intervention.

Page 3

It is highly likely that the neurotoxic effects of cumulative mercury burden,

including exposure to other sources or forms of mercury (thimerosal in products

other than vaccines; methylmercury in contaminated fish), follow similar patterns

of genetic restriction; it is also likely that similar genetic factors influence the

neurotoxicity observed following exposure to xenobiotics other than mercury

(e.g., PCBs, the PBDEs used as flame retardants in computers, and infectious

agents). Age and developmental status at the time of exposure, nutritional

factors, and gender are also known to influence outcomes. We have limited

ability to explain the interplay of such factors in humans; consider the example of

the disparate cognitive outcomes reported in children in the Faroe Islands and

the Seychelles after similar prenatal methylmercury exposures. The reasons for

this divergence remain unclear. The design of future epidemiologic studies must

take into account the possibility of multiple xenobiotic exposures as well as the

influence of factors that modulate risk. Our studies have important implications

for understanding the role of gene-environment interactions in the pathogenesis

of autism and related neurodevelopmental disorders.

I refer Subcommittee Members to our recent publication in Molecular Psychiatry

where experimental findings and their implications are discussed in more detail.

Thank you for your attention.

Mady Hornig, MD

New York, NY