Wakefield's Testimony before the Congressional Committee

TESTIMONY TO THE CONGRESSIONAL OVERSIGHT COMMITTEE ON
GOVERNMENT REFORM

Dr Andrew J Wakefield MB BS FRCS

Introduction

Mr. Chairman and members of the Committee, I am here
to review the progress in our understanding of the
possible association between childhood developmental
disorders, gastrointestinal disease and vaccines, and
to make certain recommendations.

This area of inquiry is beset with issues that go way
beyond the clinical and scientific evaluation and
treatment of affected children; implications for
Public Health policy are far reaching and demand an
early resolution to the issue of whether or not there
is a causal relationship between MMR vaccine - the
focus of my own particular inquiries - and childhood
autistic spectrum disorder (ASD).

I consider many of the associated issues, including
attacks on personal integrity, peripheral and will not
address them here. Rather, it is my wish to bring the
Committee up to date with the clinical science, and to
try and offer a positive and achievable strategy for
addressing the important issues, for the sake of
affected children and for the protection of future
generations.

I have spent the last 36 months travelling widely, and
studying the subject in some depth, in order to
broaden my own understanding of the many aspects of
the current dilemma. My starting point, as is the
starting point for much of medical science, is with
the clinical histories of the individual children.
These have been an essential element in my own
understanding of the issues. Having reviewed a large
number of these histories it is only now that I am in
a position to propose hypotheses that are capable of
being examined by epidemiological methods. Some of the
factors, relevant to the hypotheses, will be set out
later in this written testimony.

I will start by bringing the Committee up to date with
the research findings.

Progress

At the last congressional hearing, Professor J.
O'Leary and I presented results of a detailed analysis
of the clinical and pathological features, and early
virological findings, in a group of 60 children with
regressive autism.

These original data, and subsequent investigations
confirm that there is a group of children with autism
in whom there is a consistent and characteristic
pattern of pathology in the gastrointestinal tract.
This pathology consists of lymphoid nodular
hyperplasia (swelling of lymph glands) in the terminal
ileum and large intestine (colon), and inflammation in
the mucosal lining of the intestine.

The initial report describing this possible new
syndrome of regressive autism and intestinal
inflammation was published in the Lancet 1998[1]. The
data describing the first 60 children, comparing them
with appropriate controls, were published subsequently
in the American Journal of Gastroenterology [2]. We
have now investigated nearly 200 affected children,
and continue to detect the same pathological findings
in the intestine. The recognition of this syndrome is
rapidly gaining wider acceptance among
Gastroenterologists. In an editorial accompanying the
second paper2, the Editor of the American Journal of
Gastroenterology wrote:

"[the authors] are to be congratulated on opening yet
another window onto the ever-broadening spectrum of
gut-brain interactions. Their findings raise many
challenging questions that should provoke further
much-needed research in this area, research that may
provide true grounds for optimism for affected
patients and their families."[3]

Professor Eamon M. M. Quigley

Dept of Medicine, National University of Ireland, Cork


A subsequent detailed analysis of the immune
characteristics of the mucosal lesion in the large
intestine (colon) was published recently in the
Journal of Pediatrics[4]. This study confirmed the
presence of an apparently novel form of
immune-mediated inflammatory bowel disease in children
with ASD. The intestinal findings were not seen in
developmentally normal children. The data are
consistent with increasing evidence for damage and
dysfunction of the lining of the bowel (epithelium) in
this cohort of children. The paper concludes that:
"the increasing evidence of immunopathology [in
children with autism] suggests that focus on
autoimmunity, rather than genetics, may have now
become a priority."

In a highly supportive editorial comment, the Editor
of The Journal of Pediatrics concluded:

"This [the study findings] seems to point to gut
epithelial dysfunction leading to altered permeability
and subsequent entry of "CNS altering substances." It
follows that treating the gut disease may affect the
CNS disease.

Journal of Pediatrics March 2001.
Editorial Comment.- William F Balistreri MD.

The small intestinal lesion

Similar detailed investigations of the small
intestinal mucosa in these children has confirmed the
presence of an immune pathology, characterised by
infiltration of the epithelial lining by CD8+
lymphocytes, structural damage to the epithelium and
binding of serum antibodies to the epithelial
membrane.[5] [6] These observations distinguish the
children with autism from healthy paediatric controls
and those with alternative inflammatory pathology,
such as, coeliac disease. The findings are consistent
with an autoimmune disease of the intestine, and are
similar to the autoimmune intestinal disease seen in
some patients with AIDS.

Horvath and colleagues, from the University of
Maryland Medical School have published clear evidence
of upper gastrointestinal (i.e. esophagus, stomach and
duodenum) disease in the majority of autistic children
presenting with gastrointestinal symptoms [7]. Their
findings include inflammation of the esophagus,
stomach and duodenum, with associated deficiencies in
the digestive enzymes that are normally produced in
the small intestinal epithelium. Based upon similar
symptoms in children under our care, and the findings
of Horvath et al, we have included a similar
endoscopic evaluation of symptomatic children and have
confirmed the high prevalence of inflammatory
pathology in these areas, consistent with Horvath's
findings (manuscript in preparation).

The gut-brain connection

One outstanding question applies to the nature of the
link between the disease in the diseased intestine and
the developmental and behavioural problems in affected
children. A manuscript is appended to this testimony,
that elaborates the evidence for a toxic gut
brain-interaction in this subset of children,
providing a biologically plausible - and testable -
mechanism for how a primary intestinal pathology might
lead to neurological damage [8]. It is acknowledged
that in view of the immune disorders evident in these
children, there may well be an autoimmune damage to
structures within the developing brain, and this
possibility is under investigation. Toxic and
immune-mediated cerebral pathologies are by no means
mutually exclusive. Included in the appended
manuscript is reference to two further papers that
have been published in the last several years which
have a direct bearing on our growing understanding of
the role of the gut in developmental disorders,
including regressive autism. Bolte has proposed a role
for the toxic products of certain species of bacteria
in the intestine, in mediating the encephalopathy that
is associated with behavioural disturbances in
affected children [9]. This situation would be
analogous to the encephalopathy that is seen in
patients with acute or chronic liver failure, where
toxic bacterial metabolites are absorbed into the
bloodstream from the gut but not removed by the liver,
resulting in characteristic encephalopathic changes in
the patient. In seeking to test Bolte's hypothesis,
Sandler et al conducted an open-label study of oral
antibiotic (Vancomycin) therapy in children with
autism [10]. Vancomycin is an antibiotic that is not
absorbed from the gut. The authors demonstrated a
clear cognitive improvement in children during the
treatment period. Children deteriorated behaviourally
following cessation of Vancomycin therapy, suggesting
that the aberrant effects of the intestinal bacteria
were likely to be the consequence of an underlying
intestinal abnormality, as indicated by our own
studies, but that the products of these bacteria play
an important role in mediating the gut-brain
interaction and associated behavioural pathology.

In summary, there is growing evidence that toxic
products from the diet and intestinal bacteria leak
through the damaged bowel and injure the developing
brain.

Nonetheless, the studies of Bolte and Sandler provide
an important insight into a possible mechanism of
gut-brain interaction that is likely to be important
in these children with regressive autism.

To what proportion of autistic children is the finding
of gastrointestinal disease relevant? From independent
sources, both Melmed et al [11], and Horvath et al
[12] have shown that gastrointestinal symptoms are
common in children with autism compared with
developmentally normal controls. The symptoms
described by these authors mirror those in children
seen in our clinic and are, we believe, likely to
reflect identical pathology. Further studies of
autistic children who have no overt gastrointestinal
symptoms indicate underling intestinal pathology [13].
These children may possess the intestinal pathology
that affects the brain but do not demonstrate
sufficient symptoms to be clinically significant or
alert a parent to the problem. These observations are
reminiscent of celiac disease, an allergic sensitivity
to dietary gluten, where a large proportion of cases
may be subclinical, and are only detected by
population screening. This phenomenon is known as the
"celiac iceberg".

It is vitally important to recognise this phenomenon
since those with subclinical celiac disease are at
risk of long-term complications unless appropriate
therapy is instituted. It is highly likely that a
similar iceberg effect is operating in children with
autism, and that there may be either a primary, or
associated intestinal pathology in many affected
children.

Since last years' Congressional hearing a number of
meetings have taken place in the U.S. to consider the
issues. The first was June 2000 at the American
Academy of Pediatrics. Due to the unavailability of
invitees from the CDC the original meeting was
rescheduled for a date that made it impossible for
Professor O'Leary and I to attend. The deliberations
of this meeting have yet to be published.

The second significant meeting was convened by Dr. Ian
Lipkin of the Emerging Pathogens Laboratory,
University of California, Irvine. This meeting took
place February 2001 at Cold Spring Harbour and was,
for the most part, extremely constructive. The
organisers are to be commended on this initiative. At
this meeting confirmation of the detection of measles
virus in the intestine of children with ASD was
endorsed (see below). The proceedings of this meeting
are due to be published as part of a series of invited
articles in Molecular Psychiatry. It was confirmed
that independent virological studies of affected
children, focusing on the detection of measles virus
predominantly in circulating immune cells, were
planned. I strongly encouraged the testing of
intestinal biopsy samples in addition to blood
samples, since viral detection is likely to be more
successful.

The third meeting, convened as a direct consequence of
a request from this Committee, took place at the
National Academy of Sciences, Institute of Medicine in
March 2001. Presentation of the deliberations of this
Committee are due to be reported this week.
Unfortunately, due to the presence of the press,
virological data undergoing peer review, could not be
presented, which largely defeated the purpose of the
meeting.

Summary

In summary, primary intestinal pathology may be a
significant part of the disease process in a large,
but as yet, undefined proportion of the children with
autism. In the great majority of autistic children who
have been investigated appropriately, according to
their intestinal symptoms, there is inflammatory
pathology the features of which are consistent with an
autoimmune mucosal lesion in both the large and small
intestine. We propose that this intestinal disease
makes a major contribution to the
developmental/behavioural pathology in affected
children. The findings, in these ASD children, of an
autoimmune intestinal disease, associated intestinal
lymphoid hyperplasia, and immunodeficiency, are
consistent with a viral cause for this syndrome.

Recommendations

1.Based upon the peer reviewed medical and scientific
literature, children with developmental disorders and
gastrointestinal symptoms should be thoroughly
investigated for underlying intestinal pathology,
according to the principles of evidence-based
medicine. The associated intestinal pathology is
amenable to therapy and merits active treatment.

· These children deserve investigation to rule in or
rule out organic immune and intestinal pathology

· Only with investigation and diagnosis can
appropriate therapy be instituted

· Because of their behavioural diagnosis, these
children are often denied access to appropriate
healthcare. It is evident that for many children,
their intestinal and other physical symptoms are put
down to their behavioural problem, and are not being
investigated appropriately. This situation must not be
allowed to continue.

· To exclude these children from appropriately
resourced medical intervention is discriminatory and
unacceptable. Indeed, any such discriminatory policy
may actively prejudice the perception of health care
providers and make them reluctant to institute
necessary medical investigations and treatment of
these children.

2. That investigation of intestinal symptoms -
hitherto largely ignored in children with
developmental disorders - should be given high
priority, and that the identification of screening
tools to identify autistic children with subclinical
intestinal disease are given a similarly high
priority.

3. That, consistent with the IOM's recommendations
published this week, structured, multidisciplinary
gastrointestinal, immunological microbiological and
neurodevelopmental research programmes should be
initiated between collaborating institutions.

4. That, in order to institute points 1-3. above, a
high level strategic meeting be called, and a working
group established, under the auspices of the American
Gastroenterological Association, to include members of
the American professional societies for Microbiology,
Immunology, Pediatrics, Neurology and, in addition,
Autism specialty groups such as DAN (Defeat Autism
Now).

5. That this meeting should seek to define a strategy
for:

· clinical investigation and management of affected
children;

· application of our extensive knowledge in the
respective areas of inflammatory bowel disease,
intestinal motility disorders, mucosal immunology,
intestinal microbiology, nutrition, food allergy
toxicology, and neuroimaging, to the study of the
gut-brain-immune axis in affected children.

6. That the working group should strive to develop (I)
a management guide for clinicians and (2) set
priorities for clinical and basic scientific
investigation.

My colleagues and I would be happy to help co-ordinate
such a meeting and to nominate names of possible
members of the working group whose contribution I
would see as essential for its success.

Virology

The majority of parents seen in our clinic and many
parents of similarly autistic children throughout the
developed world cite exposure of the MMR vaccine as
the trigger for regression in their previously normal
child. For many reasons, not least of which are the
benefits of vaccination, clinicians and scientists
have been unwilling to entertain the possibility of a
causal relationship.

The identification of an immune mediated intestinal
pathology has provided a focus for investigation of
this possibility. Specifically, we have tested the
hypothesis that measles virus (or elements thereof)
should be detectable in the enlarged lymph nodes in
the ileum. We have now completed a preliminary series
of studies which sought to answer the following
questions:

· Is measles virus present in the intestinal tissues
of affected children and specifically, can measles
virus protein and genetic material be detected in the
same location?

· If measles is detected where is it located?
Specifically is it present in the swollen ileal lymph
nodes?

· How many viruses is there?

· Can the virus be sequenced in order to characterise
the strain of the virus present?

· Can the results be confirmed by different
technologies?

· Does the presence of the measles virus distinguish
autistic children from controls?

· Can the results be confirmed in independent
laboratories?

In a coded-blinded study using appropriate positive
and negative controls these questions have largely
been answered. The data are not presented here, since
this may prejudice the process of peer review and
publication. However, it is possible to confirm that:
· Both measles virus gene and protein are present in
intestinal tissues from a majority (93%) of autistic
children studied. A variety of different measles virus
genes were detected

· The virus is present in the reactive lymph nodes.

· The virus is present in specific cells that would
make it a likely cause for the lymph node reaction.
Other common childhood viruses including adenovirus,
herpes virus, mumps, rubella and HIV were not present.


· The virus was present in relatively small amounts,
making the application of highly sensitive molecular
detection techniques an essential component of these
studies.

· Viral sequencing confirms the presence of measles
virus. Strain-specific sequencing studies (i.e. to
discriminate wild and vaccine strains) are currently
under way.

· Overall ten different techniques have been applied
to detect measles virus in these biopsies, with all
reporting positive results

· There is a highly statistically significant
difference between cases and controls for the presence
of measles virus in ileal biopsies. It is notable that
measles virus was present in a small percentage (11%)
of biopsies from developmentally normal children.

These data are due to be presented at Digestive
Diseases Week, Atlanta, GA, May 2001.

Collaborative studies with independent laboratories
are being undertaken in order to assess the
reproducibility of these findings. An invitation has
been extended to members of the CDC to spend time in
the laboratory of Professor O'Leary in order to
participate in the analysis of tissue. In addition,
independent studies analysing in intestinal biopsies
from children in Canada and the United States have
been prompted by these investigations.

Part of my efforts over the last 24 months have been
to encourage and initiate some of these studies. At
the recent Cold Spring Harbour meeting on Autism and
Environmental Agents it was the expert's conclusion
that the evidence confirmed the presence of measles
virus in intestinal biopses from children with ASD. It
is important to note that the data identify an
association only; the presence of the virus does not
make it the cause of the autistic regression. Whether
or not this is the case will only be determined by
further study. Once again, the virology data are
consistent with, but not proof of a causal association
between measles virus and the intestinal disease in
children with ASD, and in view of the plausibility of
the gut-brain scenario described above, the data are
also consistent with - but not proof of - measles
virus causing both intestinal disease and autistic
regression in these children.

The findings move the arguments to a new level that
have implications for public health policy. Rather
than having parental reports alone, of a temporal
association between MMR exposure and developmental
regression, there is now definitive evidence of an
novel and specific pathology in the intestine of
children with ASD that is associated with the presence
of measles virus. In association with the findings of
Kawashima et al, of measles virus in the peripheral
blood of some children with ASD, it is no longer an
correct or acceptable to state that there is no
evidence of an association between MMR and this
syndrome. In light of this and in view of the
acknowledged lack of adequate safety studies on the
MMR vaccine [14] the case for making MMR vaccination
either mandatory, or the exclusive mode of protection
against measles, mumps and rubella is, in my opinion,
difficult to justify. Parents should be given an
informed choice of vaccination strategy, including the
provision of single vaccines. I fully endorse the need
to continue to protect children against these
infectious diseases.

At present we are performing strain-specific
sequencing on the virus to determine whether the virus
in the intestine of these ASD children it is a vaccine
or 'wild type' (natural) strain. The current teaching
is that there is no evidence for persistent infection
with the vaccine strain virus. It must be assumed that
this would be an undesirable consequence of
vaccination in view of the association between
persistent infection and delayed severe neurological
disease.

Vaccination is the only recorded exposure these
children have to measles, and wild measles has not
been circulating within their lifetime. However, it
could be that the vaccine failed to protect them from
(and may, in fact, have predisposed them to) a covert
atypical natural measles infection. There are both
clinical and experimental data demonstrating that
failure of a measles vaccine to produce adequate
immunity may actually predispose to immune disease
upon re-exposure to measles virus.

Alternatively, the presence of the virus may
represent innocuous carriage, although I consider it
would be imprudent to make this assumption in the face
of the immune and pathological data, prior to thorough
investigation.

Clear priorities must be, to:

· conclude the strain-specific sequencing studies,

· conduct investigations of measles-specific immune
responses in affected children and appropriate
controls.

· fund independent virological studies.

· include the development of a strategy for the
conduct of viral studies in the remit of the working
group as outlined above.

Review of MMR vaccine safety

In early 2001 Dr Montgomery and I published a critical
review of pre-licensing studies of MMR safety[15] (the
bibliography for this section is included in reference
16). In this review we raised substantial concerns
over the inadequacy of these safety trials, in terms
of the numbers of individuals studied, the lack of
appropriate control groups, and the lack of any long
term safety evaluation. In particular, we raised
concerns over the use of combination of live viruses
in the MMR vaccine. These concerns were endorsed by
the peer reviewers, whose comments were published
alongside the review.

This review has been criticised as being inaccurate
and selective in certain respects. We readily
acknowledge the omission of one substantial paper.
This was not due to deliberate selection, but due to
failure to identify the study in our literature
search. This omission was a study of twins from Sweden
receiving MMR vaccine. The study was well designed,
although it sought to detect adverse events occurring
only within three weeks of vaccination. Despite its
omission, this study involved far too short a period
for the adequate evaluation of safety, and we do not
believe that it mitigates against the concerns raised
in our review. We reject other criticisms as
inaccurate and misrepresentative of the substance of
our review[16].

Viral "interference"

In the review, particular concerns were raised in
relation to the potential for "interference" between
the component viruses of MMR. This phenomenon, which
is universally recognised by virologists, refers to
the influence of one virus upon another when they are
encountered by the host concurrently, or in close
temporal sequence. The way in which one virus
interferes with another is not fully understood;
however, a substantial component of this interference
is through an influence of one virus upon the immune
systems' ability to deal effectively with another
virus. In human disease viral interference may
profoundly and adversely affect the nature and outcome
of a patient's disease.

Therefore, a major safety consideration in the context
of polyvalent vaccines such as MMR, is the potential
for adverse interactions between the component live
viruses, particularly in view of the immunosuppressive
properties of measles virus. In addition to the
elements of unnatural age, route, dose, and strain of
infectious exposure, the childhood immune system must
cope with a combination of viruses that it would have
been extremely unlikely to encounter under
circumstances of natural exposure. In an executive
summary, members of the IOM's Vaccine Safety
Committee, reiterated this anxiety in the context of
virus-induced immunosuppression and polyvalent
vaccines. They stated:

"It may be asked, then, whether the use of combination
viral vaccines might exacerbate the potential problem
of immune suppression. The committee found no report
of a systematic comparison of the effects of
monovalent and polyvalent live attenuated vaccines on
immunity".

In 1995 concerns over the potential for interference
between the components of vaccines were raised again
at a meeting of US vaccine officials. Specifically,
Belshe (St Louis) stated that:

"To be confident that a particular vaccine had no
effect on another vaccine given simultaneously,
comparative studies should be performed".

Halsey (Johns Hopkins) considered that such studies
would be both "too large" and "unnecessary". Halsey
conceded, however, that: "If there is a biological
reason to suspect that there may be interference or
blunting or blocking, then comparative studies should
be done."

Is there a "biological reason" to suspect that
"interference" may occur between the component viruses
of MMR? It is evident from the medical literature
prior to 1977, that the outcome from measles infection
may be influenced by close temporal exposure to
another virus. A close temporal exposure to measles
virus and another infection, for example, chickenpox
or certain enterovirus infections, is associated with
an excess risk for delayed encephalitis. With respect
to possible adverse events that are currently topical,
atypical patterns of exposure to measles, mumps,
rubella and chickenpox have been associated with both
autism and, for measles virus, developmental
regression. In utero and infant exposures have been
identified as periods of apparent susceptibility, when
both the brain and the immune system are undergoing
rapid development. It is notable that a close temporal
relationship in the exposure to more than one of these
infections during periods of susceptibility, may
compound both the risk and severity of autism.
Similarly, atypical patterns of measles infection,
including a close temporal exposure to mumps
infection, but not other common childhood infections,
have been identified as a significant risk factor for
classical inflammatory bowel disease, Crohn's disease
and ulcerative colitis.

Clues that the component viruses of MMR could
interfere, one with another, were provided in the very
first pilot studies of this vaccine. In 1969, Buynak
et al sought to examine the effects, in humans, of
various combinations of measles, mumps and rubella
strains. In addition to seroconversion (viral antibody
production), clinical end-points included the
comparative frequency of measles rash and fever.

The authors demonstrated clear evidence of dose- and
strain-dependent interference between the component
viruses in the MMR vaccine. However, despite the
potential implications for safety the matter was not
followed up by these authors.

Six years after Buynak's study, in 1974, the potential
for interference in MMR was subject to a more detailed
follow up of the original observations, by Minekawa et
al. The most striking observation was of a
dose-dependent influence of the mumps vaccine upon not
only clinical reactions to the measles component, but
also seroconversion to rubella vaccine. This same
pattern of interference was also indicated by the
study of Eddes et al that compared clinical reactions
to monovalent measles and MMR vaccines.

The ability of mumps virus to interfere with the
cellular immune response to measles virus and,
thereby, to potentially impair viral clearance and
increase the risk of persistent infection and/or
initiate autoimmune disease, is a real and worrying
possibiilty to some of those involved in the current
debate. The contemporaneous interpretation of Minekawa
et al was that further studies were necessary. From
the published literature it does not appear that any
further studies were undertaken.

Further evidence of viral interference – in this
instance, between the measles and rubella vaccines –
comes from Crawford and Gremillion's study of U.S.
Airforce recruits in 1981. In a relatively large
prospective study, safety and efficacy of measles and
rubella vaccines (given either alone or in
combination) were compared with unvaccinated controls.
Five hundred and twelve vaccinees were compared with
835 unvaccinated controls and data were stratified by
sex. The authors noted an increase in reports of fever
and diarrhoea in those immunized with both vaccines
simultaneously. In women there was an increase in
complaints of myalgia (muscle pain) after simultaneous
immunization. The data merit more detailed
consideration; in recruits receiving either monovalent
measles or rubella vaccines there was no significant
increase in diarrhoea compared with unvaccinated
controls (measles vaccinees versus controls [men] OR
2.51; CI 0.06-9.99) and [women] OR 3.61; CI
0.26-50.42; p 0.5; Odds ratios for rubella vaccinees
versus controls cannot be calculated since no men or
women reported diarrhoea after rubella vaccine alone.
In contrast, compared with unvaccinated controls there
was a significantly increased risk of diarrhoea
following simultaneous measles and rubella vaccination
in both men (OR 7.31; CI 1.11-34.64) p<0.001) and
women (OR 17.29; CI 1.14-247.09; p<0.001). For
gastrointestinal adverse events (diarrhoea), the
effect of simultaneous measles and rubella vaccination
is not additive but apparently synergistic (compound).
These findings were obvious to the authors and were
remarked upon in the results. In the jargon of vaccine
regulators, a signal had been generated, but it
appears to have been ignored.

Indications that novel adverse events might be
associated with the combined MMR vaccine, rather than
the monovalent component vaccines, have come from
Plesner et al's study of gait disturbance following
MMR in Denmark. Several prior studies had indicated
that gait disturbance (ataxia) might occur in up to 1
in 1000-4000 recipients of MMR. In Denmark this
association had not been detected with any other
vaccine administered to children of the same age,
prior to the introduction of MMR in 1987. In a recent
follow up of the mandatory passive reporting system
operated in Denmark, Plesner not only confirmed this
association but also indicated that the more severe
ataxias following MMR may be associated with residual
cognitive deficits in some children. This association
is specifically relevant to the debate on MMR and
autism, as parents of autistic children who suspect a
link with MMR, not infrequently report gait
disturbances.

Vaccine manufacturers recognize that the problem of
interference exists, but appear to have regarded it as
more of an inconvenience that a safety concern.
Douglas of Merck stated recently:

"The complexity of vaccine delivery today in clinical
practice with 15-17 injections in the first two years
of life emphasizes the need for development of
combination pediatric vaccines, for example, putting
DTaP, HBV, HIB and IPV together. This has proved to be
far more difficult than previously believed due to
unpredicted immune interference and incompatibilities
on mixing of different components, demonstrating again
the inadequacy of our understanding of how vaccines
work and the empiric nature of the science.

Why, however, in spite of evidence provided by studies
undertaken two decades earlier should such
interference be considered "unpredictable" and,
indeed, remain unstudied?

In summary, the data indicate that the combined MMR
may be associated with novel adverse events (i.e. not
seen with the single vaccines) including regressive
autism. These novel adverse events may arise because
of the interaction of the viruses (with each other
and/or with the immune system) that leads to an
increased risk of autoimmune disease. Knowledge of
viral interference in MMR has been available for over
30 years but has never been adequately studied.

Shortcomings in current epidemiology

Currently available epidemiological studies are
inadequate to rule in or rule out a causal
relationship between MMR and ASD. Several recent
epidemiological studies have sought to examine this
association [17] [18]. Description of the details of
these studies will be left to expert witnesses. These
studies have involved the analysis of time trends in
the incidence22 or case load21 of autism in defined
populations (UK and California). These studies sought
to examine whether there was any correlation between
the dramatic increase in new cases of autism (which
was confirmed) and MMR vaccine uptake. In their
analyses no correlation was found. These studies have
been portrayed by some as testing the 'Wakefield
Hypothesis'. Categorically, this is not the case and
any such interpretation would be wholly inappropriate.


I will attempt to clarify this statement. The work of
my group and our collaborators colleagues has been
concerned with the identification and description of
an apparently novel disease complex in children. In
the context of this disease we have not yet described
or proposed any hypothesis that is testable in an
epidemiological context. We are only just beginning to
understand the complex nature and clinical
characteristics of the syndrome. As such, we are not
yet in a position to provide a comprehensive
case-definition. Of particular relevance to the design
of epidemiological studies, we have been concerned to
identify any vulnerability factors that might put a
child at increased risk of developmental regression
following MMR vaccination.

My colleagues and I have now taken detailed clinical
histories from parents of over 250 affected children
in our own clinic, and I have traveled extensively in
the last 3 years for the purpose of listening to
clinical histories from parents and health care
providers across the United States, Canada, and Europe
in order to assimilate these histories and to compare
them with our own experience.

In particular, the clinical histories have identified
the following circumstances that are frequently
associated with the receipt of the MMR vaccine, by
parents of affected children. These include MMR
vaccination of a child who:

· has an current infection (e.g. ear or upper
respiratory tract infection

· is either receiving antibiotics, or is vaccinated
shortly after a course of antibiotics;

· has a history of atopy (allergy), particularly milk
allergy;

· has received multiple vaccines concurrently;

· has a strong family history of autoimmune disease;
and,

· in light of the critical observations reported by Dr
F. Yazbak, a maternal history of MMR/rubella
vaccination immediately prior to pregnancy, during
pregnancy, or during the period of post-partum
nursing.

These clinical associations may represent co-factors
for, or makers of, a susceptibility to an adverse
reaction to MMR vaccine, and specifically the syndrome
described above. The identification of these potential
co-factors allows us to predict what the effect of the
combination of MMR plus such co-factors would be, over
time, at the population level, if MMR were causally
related to ASD. We would hypothesise that:

· following the introduction of MMR there would be an
increase in the incidence of autism from previous
baseline levels;

· that the background vulnerability for ASD following
MMR would increase in successive birth cohorts
(groups) as the rate exposure to these co-factors -
infant allergy, maternal autoimmune disease, infant
antibiotic use and additional vaccine (including the
associated preservatives and adjuvants) administration
- increased over time. The increase in all of these
potential co-factors has been clearly documented. The
issue of antibiotics and novel vaccines is
particularly important in this respect, since these
may exert potent influences on the developing immune
system and it is biologically plausible that they may
render the infant less able to handle a live viral
vaccine given either at the same time or subsequently.


· That, in view of the high uptake of MMR and the
concomitant increase over the last 2 decades in infant
exposure to potential co-factors, the incidence of
autism would continue to rise beyond the introduction
and widespread use of the MMR vaccine.

It is important to note that this hypothesis has been
arrived at independently, following a detailed
examination of the clinical histories of affected
children, and an assessment of the biological
plausibility of the sequence of immune and
pathological events. It has not been influenced by the
publications of either Kaye et al20 or Dales at al21,
although their findings are consistent with, and
strongly supportive of, this hypothesis. It is my
impression that the trend in autism incidence is
wholly consistent with a causal relationship with the
MMR vaccine at the population level.

For the sake of completeness I have included, in
non-lay terminology a summary of the current
hypothesis for the origins of the syndrome in this
cohort of children with autism. Hypothesis For the
cohort of children with regressive autism – a group
that, we believe, may account for a large proportion
of new diagnoses – we hypothesize that the root
problem is in aberrant early immune programming,
particularly within the mucosal immune system. Within
the last decade there has been substantial increase in
allergies of all kinds, particularly dietary, which
may result from recent dramatic changes in the pattern
of infant environmental exposures. There are
increasingly explicit links between mucosal infectious
exposures and the establishment and maintenance of
mucosal immune tolerance. The natural trajectory over
time, from a TH2 dominant foetal/neonatal immune
response to a balanced TH2/TH1 responsiveness that
likely reflects healthy immunological maturity, may
permit the generation of appropriate cytotoxicity in
the face of viral exposures. Factors that modify this
transition in T helper cell effector function,
including vaccines, toxins or natural infections, may
prolong TH2 skewing, and thus impair antiviral
responses. Inappropriate early conditioning of the
mucosal immune system, for which the faecal flora
plays an obligatory role, may allow inappropriate
persistence of agents, which home to gut-associated
lymphoid tissue. The immunomodulatory nature of MV
suggests that persistent expression within mucosal
lymphoid tissue may affect mucosal tolerance
mechanisms, in particular inducing TH2-skewing through
mechanisms including inhibition of dendritic cell
IL-12 production.

If it is the case that MMR vaccine is causally related
to this syndrome – a case as yet unproven - it should
not be assumed that the associated risk remains static
within any given population over time. The rapid
increase in numbers of children with dietary allergy,
itself associated with reduced CD8 cell numbers,
prolonged viral infections and familial autoimmunity,
suggests that the numbers of children who may be at
risk of aberrant responses to infectious agents will
have risen in the last decades. Potentially relevant
overlap, in which ubiquitous infectious exposure is
followed in a few children by autoimmune-mediated
neuropsychiatric abnormality, occurs in the paediatric
autoimmune neuropsychiatric disorders associated with
streptococcal infection (PANDAS). Population-based
epidemiological studies not have been helpful in the
identification of this association. The likely
autoimmune basis of regressive autism suggests that
any causal association with MMR vaccine would lead to
a continuing upward trend in incidence after vaccine
introduction - in developed-world but not
developing-world populations, in parallel with other
autoimmune diseases – rather than the reported
epidemiological construct of a step-up-and-plateau
model.

Recommendations

· That virological studies of children with ASD,
including appropriately designed large-scale
epidemiology, are made a priority and that the
strategy for these investigations be set out by the
working group, as described above.

· That immediate active surveillance of vaccine
related adverse events be instituted, and that the
responsibility for this surveillance be independent of
those authorities either promoting vaccines, or
involved in the manufacture or licensing of vaccines.


Conflicts of interest

I am the named inventor on a viral diagnostics patent
in the area of Crohn’s disease and ulcerative colitis.
I derive no personal income or other financial
benefits from this patent. Any future revenues from
these patents will be used to fund our medical
research programme into inflammatory diseases of the
gastrointestinal tract.

I am acting as an medical expert in the current MMR
class action in the UK

I hold no stock in any company and I am not in receipt
of any grants from vaccine manufacturers.

--------------------------------------------------------------------------------

[1] Wakefield, A.J., et al Lancet 1998; 351:673-641
[2] Wakefield, A.J., et al Am J Gastroenterol
2000;95:2285-2295
[3] Quigley, M.M., Am J Gastroenterol
2000;95:2154-2145
[4] Furlano, R. et al., Journal of Pediatrics
2001;138:366-372
[5] Torrente, F. et al., Journal of Paediatrics,
Gastroenterology and Nutrition. 2000; 31 (Suppl 2)
A546 (manuscript in preparation)
[6] Wakefield, A.J. & Murch, S.H., Molecular
Psychiatry (In press)
[7] Horvath, K. et al., Journal of Pediatrics
1999;135:559-563
[8] Wakefield, A.J.et al., Entero-colonic
encephalopathy, autsim and opioid receptor ligands (in
preparation)
[9] Bolte, E.J., Medical Hypotheses 1998; 51:133-144
[10] Sandler, R. et al., Journal of Child Neurology
2000; 15:429-435
[11] Melmed, R. et al., Journal of Paediatric
Gastroenterology and Nutrition. 2000;31 (suppl 2) A116
[12] Horvath, K. et al., (in press)
[13] D'Eufemia et al., Acta Paedtatrica
1996;85:1076-1079
[14] Grune and Stratton IOM
[15] Wakefield, A.J. and Montgomery, S.H., Adverse
Drug reactions & Toxicology Review. 2000; 19:265-283
[16] For example, we presented a re-analyses of the
adverse reactions data in the original studies of MMR
by Stokes 1971. Our paper was concerned with issues of
safety and we presented a reanalysis of the safety
data. Stokes' paper quotes vaccine administration to
larger numbers of children for whom safety data are
not provided and for whom, therefore, reanalysis was
not possible.
[17] Kaye, J. A. et al., BMJ2001;322:0-2
[18] Dales, L. et al., JAMA2001;285:1183-1185