June 1999:
Report of the working party on MMR
vaccine
1. Background
On the basis of their
research findings, doctors at the
Royal Free Hospital led by Dr. Andrew
Wakefield have suggested that both
measles and measles vaccines may be
causally associated with Crohn's
disease (a serious form of
inflammatory bowel disease) and MMR
vaccine with the development of
autism (a developmental disorder
usually, diagnosed in the second year
of life). MMR vaccine was introduced
into routine UK immunisation
programmes in 1988, following which
it has been given to millions of
children at the age of 13-15 months.
1.1 Sources of
information
During 1996-7 the MHRA
(medicines) (MHRA) was informed by a
firm of solicitors that they had
received several hundred reports of
children who had developed autism,
Crohn's disease or similar disorders
after immunisation with measles,
mumps and rubella (MMR) vaccine or,
much less frequently, measles rubella
(MR) vaccine. Such reports were
continuing to accumulate and a wide
variety of other conditions had also
been reported.
In order that more
information could be obtained about
these possible adverse effects,
purpose-designed questionnaires were
sent to parents via the solicitors.
To validate information received from
parents, further questionnaires were
sent to the doctors (the GP and at
least one specialist) who had cared
for these children.
1.2 Formation
and remit of Working Party
In 1998, on the advice
of the Committee on Safety of
Medicines, an ad hoc Working Party
was formed to assess these parent
reports together with medical
evidence received from GPs and
specialists. The Working Party
included members with specialist
expertise in the fields of adult and
paediatric gastroenterology, general
paediatrics, paediatric neurology and
child psychiatry. Its membership and
remit are given at Annex 1.
1.3 Meetings of
the Working Party
The first meeting took
place on 27 February 1998. At this
meeting members of the Working Party
recognised that there were marked
limitations in studying the selected
group of children that had been
identified by the firm of solicitors
involved. The children had been
brought to the notice of the
solicitors because of a reported
association between the
administration of MMR or MR vaccine
and the appearance of certain
symptoms and signs. They were not a
randomly selected or population-based
group and there were no controls.
Because of the way in which children
had been selected, it seemed likely
that there would be a temporal
association between administration of
the vaccine and the appearance of
symptoms in the group, and it would
be difficult to draw conclusions
about that association. However, in
view of the concerns that had been
raised about the children identified
by the solicitors, the Working Party
considered that it was appropriate to
evaluate the information that was
available about these children.
In particular, it was
important to assess whether or not
there was information in medical
records to support the parental
concerns, to look for evidence of
unusual features suggestive of a
novel syndrome and to see whether
there was any evidence of acute
neurological events associated with
immunisation which might have
predisposed children to subsequent
autism. In searching for features
that might indicate a link with
immunisation, the Working Party
agreed to systematically examine
evidence relating to medical
confirmation of the diagnosis,
temporal association between
immunisation and onset, prior history
and potential alternative causes. In
doing so it recognised that, even
when these criteria were met, they
would not necessarily be sufficient
to define a causal association.
The Working Party
considered the diagnostic criteria
which should be used for autism,
pervasive developmental disorder
(PDD), developmental delay,
encephalitis, different types of
inflammatory bowel disease and other
types of bowel disorder. Also, they
advised on the drafting of the
evaluation form which would be used
to collate the information provided
by the parents and doctors for an
individual case. At the first meeting
it was decided that a pilot study
using a sample of 20 reports of all
types of reactions would be performed
to develop the evaluation form. These
were chosen at random from the first
90 reports containing responses from
both GPs and specialists.
The second meeting of
the Working Party took place on 1
July 1998 when the results of the
pilot study were discussed. Members
amended the assessment form to
improve precision. It was noted that
adequate supporting information was
frequently not available, either
because details had not been entered
on questionnaires or because the
information was not available at
source. By July 1998 information from
180 cases had been received for which
parents suspected autism as a
possible adverse effect, of which 52
also had a gastrointestinal disorder
mentioned as an additional possible
adverse effect to MMR vaccine. It was
agreed that further reports, not
included in the pilot study, should
be examined. Each report was to be
evaluated independently by an
individual member of the Working
Party and an MHRA physician using the
revised assessment form.
This set should include
a large sample of cases recorded as
autism without bowel disease (fifty
were selected, which with the 5 cases
from the pilot study represented over
40% of such cases available at the
time), all the cases with possible
autism associated with bowel disease
and all possible cases of Crohn's
disease for which information from
both GP and specialist had been
received by 1 August 1998.
The third meeting of the
Working Party took place on 5
November 1998. At this meeting
differences of view between members
and the MHRA evaluator regarding
individual cases were resolved. The
Working Party also considered an
overall assessment of the cases of
autism and Crohn's disease which had
been evaluated in detail, considered
what further steps were appropriate
and began to formulate its
conclusions. Subsequently a draft
report was prepared and circulated to
members for comments.
At its fourth meeting on
5 March 1999 the Working Party
discussed in detail a draft report of
its findings. A fifth meeting was
held on 10 May 1999 to finalise the
report.
2. Methods of
investigation
Blank questionnaires
were provided to the firm of
solicitors who distributed them to
parents. Completed parent
questionnaires were normally returned
directly to MHRA. They were
classified and entered onto a
specially designated area of the
ADROIT computerised database (see
Figure 1). This system was
specifically designed for the purpose
of handling large numbers of
individual case reports. Medical
questionnaires were subsequently sent
to GPs and specialists. Because of
the importance of supporting medical
information for evaluation, this
process was restricted to cases for
whom medical information was received
from both these sources. The cases
were initially distributed equally
and randomly amongst experts on the
Working Party (one per case) but
subsequently, where necessary, the
evidence for each diagnosis was
further reviewed by a member with
relevant specialist expertise. A
single MHRA physician also reviewed
all the cases. The two evaluators
initially worked independently of
each other using the evaluation forms
developed for this purpose.
The Working Group
focused its attention on three groups
of children i.e. those with autism,
those with Crohn's disease and those
who had both autism and a
gastrointestinal disorder. Fifty
reports of autism alone (in addition
to five included in the pilot study),
all the reports of Crohn's disease,
and all cases of autism associated
with any gastrointestinal disorder
were evaluated in detail. Members of
the Working Group were sent all the
information available from these
reports. Where a selection had to be
made (as for the possible cases of
autism alone), this was done on the
basis of earlier receipt of medical
follow-up information.
The conclusions of the
Working Party member and MHRA
evaluator were summarised in a table
at the end of each assessment form in
the following four categories :
A.
Have either the diagnosis or
clinically relevant signs and
symptoms been confirmed medically ?
YES (Y) NO (N) UNSURE (U)
B.
Was the onset of the possible adverse
effect within six weeks of
immunisation with MMR ?
Parent=YES (Y) NO (N) NOT KNOWN (NK)
Health Professional=YES (Y) NO (N)
NOT KNOWN (NK)
C.
Was there history prior to
immunisation relevant to the possible
adverse effect?
YES (Y) NO (N) UNSURE (U)
D.
Was there evidence of other causes
for the possible adverse effect ?
YES (Y) NO (N) UNSURE (U)
Six weeks after
immunisation was chosen as a cut-off
point for a close temporal
association because this is the
maximum period in which viral
replication can be detected after
immunisation [1]. This cut-off point
has been used previously in
epidemiological studies [2,3]. No
minimum time period after vaccination
was used, although recognising that
cases would be included with
histories of symptom onset interval
too short to be likely to be
vaccine-related. In respect of the
timing of onset of each adverse
effect, the parental information took
priority over the doctor's
(reflecting normal clinical practice
because the doctor would normally
rely on the parent's account of
symptom onset).
Each report was
subsequently categorised with the
letters Y or N for each of the four
categories above. Those cases which
at least one evaluator judged
fulfilled the following specific
criteria: (a) medical confirmation of
diagnosis (b) onset within 6 weeks
following immunisation (c) no
relevant prior history (d) absence of
an alternative cause, i.e. those that
were categorised as YYNN, were
reviewed at a meeting of the Working
Party in order to reach an agreed
evaluation. As indicated above, such
a categorisation focuses on more
immediate events but is insufficient
to define a causal association.
For all other cases,
differences between the evaluations
of the Working Party member and the
MHRA physician were resolved between
them.
Medical
confirmation of the diagnoses of
autism/PDD and Crohn's disease
The diagnostic criteria
used for pervasive developmental
disorder (PDD) and autism are given
in Annex 2. Autism is a more specific
diagnosis than PDD. Although parental
reports invariably mentioned autism
rather than PDD, evidence to support
the more specific diagnosis was often
lacking.
Since fulfillment of the
diagnostic criteria for autism was
more likely to reflect the nature of
the information available than true
differences, this report presents the
evidence for both categories combined
i.e. autism/PDD. Statements relating
to "autism" have been
retained where they reflect parental
reports rather than medical
diagnoses.
If the criteria for PDD
or autism were met, or the evidence
supported Crohn's disease for the
possible adverse effect, this led to
a categorisation of YES for "Has
the diagnosis been confirmed
medically?" If the evaluator was
uncertain about the diagnosis then
UNSURE was categorised. If the
evidence did not support the
diagnosis then NO was categorised.
The evaluator was also
asked to indicate on the assessment
form if there were any extraordinary
features about each individual case.
Classification
of onset of possible adverse effect
Two values for time of
onset for signs and symptoms
according to the parent and doctor
were available for evaluation. If
either the parent or doctor indicated
that the onset of the possible
adverse effect was at any time within
6 weeks following immunisation then
this was classified as YES. If both
the parent and doctor indicated that
the onset of the possible adverse
effect was later than six weeks than
this was categorised as a NO. If
neither the parent nor doctor
indicated a time of onset for the
possible adverse effect, then this
was categorised as NOT KNOWN. The
evaluators exercised clinical
judgement in deciding whether
reported symptoms and signs were
relevant to the possible adverse
effect. For example, in a case of
autism, behavioural symptoms soon
after vaccination would have been
considered relevant but non-specific
symptoms such as pyrexia alone would
not have led to a classification of
YES.
History relevant
to possible effect prior to
immunisation
For children who had
signs and symptoms of the possible
adverse effect which were usually not
recognised as such by the parent, but
were noted by the health
professional, a categorisation of YES
was used when the reporting doctor
was confident about this evidence and
UNSURE when there was doubt. NO was
chosen to indicate no relevant past
history had been reported by both the
child's parents and doctors.
Evidence of
other causes for main possible
adverse effect
Evidence of other causes
applied in particular to children
with autism/PDD. The main evidence
was that of a first generation (i.e.
parents or siblings) family history
of speech or behavioural disorders
which might indicate a genetic
component to the adverse effect
[4,5]. More distant family history
only was not considered evidence.
Evaluation of
obstetric history
Obstetric history was
primarily sought to see if there was
evidence of neonatal encephalopathy
which might also be an alternative
cause. As details confirming a normal
obstetric history were needed, it was
not considered adequate evidence of a
normal obstetric history for a doctor
just to circle YES or NO on the
questionnaire without supporting
information. An obstetric history was
considered adequate only if the
child's doctor(s) had provided
details.
Evaluation of
signs and symptoms of encephalitis
and encephalopathy
Both the Working Group
and MHRA evaluators were provided
with checklists of signs, symptoms
and investigations for serious acute
neurological events which might
indicate encephalitis or
encephalopathy. If either condition
occurred, this might be a relevant
cause of autism/PDD.
Inclusion of
reports for evaluation
By mid-1998 about 1200
parent questionnaires had been
distributed by solicitors. On 21
October 1998, 531 parent reports had
been received and added to the ADROIT
database. Of the 531 children, 231
were classified as having autism
(with or without gastrointestinal
disorders) and 20 classified as
Crohn's disease (one of whom also had
autism). Medical information had been
received from GPs and specialists for
218 cases in total. The Working Party
considered it was not feasible to
review all the less commonly reported
effects because the body of
information available for the
individual problems was insufficient
to permit meaningful analysis. In
reaching this conclusion, the Working
Party took account of the fact that
all the less frequently reported
possible effects were of symptoms,
signs or illnesses which are
encountered naturally in children in
this age group.
On the basis described
in section 1.3 above, one hundred and
twenty-six reports were evaluated in
detail by a member of the Working
Party and the MHRA physician, as
follows: autism only (55); autism and
gastrointestinal disease (52) and
Crohn's disease only (19). Twelve
reports of autism with
gastrointestinal disorder and 3 of
Crohn's disease had to be excluded
from the full analysis because there
was insufficient medical information
available to make an evaluation.
During the evaluation it emerged that
one of the cases of "autism
only" did have gastrointestinal
symptoms. Thus, a total of 95 reports
of children with autism were fully
evaluated, of whom 54 had no
gastrointestinal problem and 41 had
both autism and a gastrointestinal
disorder. In addition, 16 reports of
children who had Crohn's disease
reported alone were fully evaluated.
Whilst all information received from
both parents and doctors relating to
reports not selected for detailed
evaluation has been entered onto the
ADROIT database, the Working Party
has not reviewed these cases
individually.
3. Results
3.1
Autism/PDD
3.1.1
Confirmation of diagnosis
Of the 95 cases of
reported autism evaluated, the
diagnosis was considered confirmed as
autism/PDD in 81, possible in 11 and
not confirmed in 3. The three
unconfirmed reports were all
considered to be cases of global
developmental delay. These three
reports are not considered in the
summary statistics below. The 92
reports of autism/PDD considered
confirmed or possible included 78
(85%) males and 14 (15%) females and
are summarised below.
3.1.2
Immunisation history
All but one of the 92
children with confirmed or possible
autism/PDD had received MMR vaccine,
the exception having received
monocomponent measles vaccine. Using
information relating to the date of
immunisation (Urabe strain vaccines
were withdrawn in the UK in September
1992) and batch numbers, it was
determined that the mumps component
of the vaccine was Jeryl-Lynn strain
for 59 children, Urabe strain for 11
and uncertain or unknown in 21.
3.1.3 Onset of
symptoms of autism in relation to
immunisation
Information on the
timing of onset of autism in relation
to immunisation as provided by
parents and doctors is summarised in
Figure 2. Some 42% of parents
reported onset of first symptoms
within six weeks following
immunisation, 47% reported onset more
than 6 weeks following immunisation
and 11% did not indicate the time of
onset. In only 6% of cases were
reporting health professionals made
aware of the child's symptoms for the
first time within six weeks following
immunisation. For 37% of reports, the
time of onset quoted by parents and
the time a health professional became
aware were similar.
Comparison of the
children whose onset of autism was
within 6 weeks following immunisation
with those for whom it began later
showed no differences in respect of
the proportions who had prior
abnormalities, relevant family
history and a confirmed diagnosis.
3.1.4 Previous
medical history suggesting
behavioural abnormalities
Doctors reported a
previous behavioural abnormality in
24 (26%), and a possible behavioural
abnormality in a further 12 (13%)
children prior to immunisation.
However, parents had only indicated a
possible pre-existing behavioural
problem in one child.
3.1.5 Family
history
In total, a family
history of speech and/or behavioural
difficulties was reported by health
professionals for 42 (46%) of cases.
For 28 children (30%), this included
at least one first-generation
relative (i.e. parent or sibling), as
follows:
Sibling only (in 5
cases it was a twin) 12
Parent only 5
Mother, uncle and brother 1
Mother and brother 1
Father and uncle 1
Father and cousin 1
Father and grandmother 1
Both parents and brother 1
Father, aunt and cousin 1
Father and aunt 1
Brother and cousin 1
Brother and multiple family
members 1
Father and multiple family
members 1
3.1.6 Obstetric
history
Adequate information to
be confident that the mother's
obstetric history was normal was
presented in only 23 reports (25%).
For 7 reports (8%), complications of
pregnancy were reported, as follows:
cord around neck (3), fetal distress
requiring Caesarean section (2),
varicella in pregnancy (1), and a
twin born second as a breech with
Apgar scores of 5 at 2 minutes and 9
at 5 minutes (1). However, none of
these was considered as evidence of
another cause for autism/PDD.
3.1.7 Signs and
symptoms of encephalopathy
One child with possible
autism was reported by a parent to
have had a convulsion within 6 weeks
after immunisation but no medical
confirmation was provided. For none
of the other cases of autism/PDD did
the parent or doctor report any
clinical signs or symptoms suggesting
a diagnosis of encephalopathy or
encephalitis. One child developed
autism/PDD concurrently with
myoclonic seizures with an onset
greater than six weeks following
immunisation.
3.1.8 Other
potentially relevant history
Three children had head
circumferences larger than the 97th
percentile. Six children had viral
illnesses which either the parent or
the doctor thought might be relevant
to developing PDD or autism, as
follows: unspecified viral illness
(3), bronchiolitis due to respiratory
syncytial virus (1), rubella (1),
measles (1). One parent had informed
the doctor that the child had started
to lose skills after a minor head
injury pre-dating immunisation.
3.1.9 Reports
initially classified as YYNN by at
least one evaluator
There were 18 reports of
autism/PDD where at least one of the
evaluators initially classified a
report as YYNN (which means confirmed
diagnosis of PDD, symptom onset
according to parents within six weeks
following immunisation, no relevant
prior history and no evidence for
alternative causes). However, in most
reports there were factors identified
which raised some doubts about either
the diagnosis, presence of preceding
speech problems, time of onset or
other causes.
For six of these 18
reports there was complete agreement
between the evaluators. After
discussion of the 12 other cases by
the Working Party a further two cases
were considered to be YYNN. None of
the eight cases considered to be YYNN
had extraordinary features suggestive
of a novel syndrome. Of the 10 cases
that were not ultimately considered
to be YYNN, doubt related to the time
of onset in 5, relevant prior history
in 3 and the diagnosis not being
confirmed in 2.
3.1.10
Interobserver variation
For all 92 reports of
confirmed or possible autism/PDD,
there was complete agreement on all
four categories between the external
and MHRA evaluators for 22 reports
(24%). Of the 70 reports where there
was disagreement, this was for one
category in 36 (39%), two categories
in 21 (23%), three categories in 9
(10%) and four categories in 4 (4%).
Differences arose over the certainty
of the diagnosis in 21 cases, time of
onset (28), presence of relevant
prior history (37), and existence of
alternative causes (35).
3.2 Cases of
Crohn's disease without autism
Sufficient information
to make an evaluation was available
for 16 reports of Crohn's disease.
3.2.1
Confirmation of diagnosis
The diagnosis was
supported by adequate information in
12 of the 16 cases. In three cases
the diagnosis was considered
possible, as follows: histological
findings consistent with but not
diagnostic of Crohn's disease (1); no
histological evidence but the child
had an ileostomy (1); orofacial
granulomatosis proven but no bowel
histology (1). For one case, Crohn's
disease was considered unconfirmed,
the available evidence being
consistent with a diagnosis of
constipation. The following
information summarises the 15 reports
considered confirmed or possible.
3.2.2
Immunisation history
For the 15 children
where Crohn's disease was considered
confirmed or possible, the parents
suspected MR vaccine in 9 cases and
MMR vaccine in 6 cases. Of the 9
children where MR vaccine was
suspected, 6 had previously received
MMR vaccine, 1 had previously
received measles vaccine, 1 had
previously received both measles and
MMR vaccine, and 1 had previously
received measles vaccine and rubella
vaccine separately.
3.2.3 Onset of
symptoms of Crohn's disease
Information on timing of
the onset of Crohn's disease as
provided by parents and doctors is
summarised in Figure 3. For five of
the 15 reports the onset according to
the parent was within six weeks of
immunisation with MMR vaccine. One
child developed arthritis between six
weeks and six months following
immunisation although bowel symptoms
started over six months later.
3.2.4
Classification of cases of Crohn's
disease
There were 4 reports of
Crohn's disease where both the MHRA
evaluator and the Working Group
member classified a report as YYNN.
None of these cases had extraordinary
features suggestive of a novel
syndrome. For all 15 reports of
confirmed or possible Crohn's
disease, there was complete agreement
on all four categories between the
external and MHRA evaluators for 10
reports. Of the 5 reports where there
was disagreement, this was for one
category in 2, two categories in 2,
three categories in none and four
categories in 1. Differences arose
over the certainty of the diagnosis
in 2 cases, time of onset (2),
presence of relevant prior history
(4), and existence of alternative
causes (2).
3.3 Other
bowel disorders associated with cases
of autism
The single case reported
on the parent questionnaire as having
both autism and Crohn's disease was
considered to have a confirmed
diagnosis of autism but not of
Crohn's disease, the medical evidence
being consistent with a diagnosis of
constipation. Thus there was no case
for which both disorders were
confirmed. Of the 41cases of autism
with a bowel disorder, the following
gastrointestinal diagnoses were
confirmed: lymphonodular hyperplasia
(9 cases: 4 alone, 4 with microscopic
colitis and one with giardiasis),
constipation (5), diarrhoea of
unknown cause (2), gastroenteritis
(2), clostridium difficile diarrhoea
(1), cryptosporidiosis (1), recurrent
intussusception (1), lactose
intolerance (1) and coeliac disease
(1). In all but two cases the onset
of the gastrointestinal disorder was
uncertain or greater than 6 months
after immunisation with MMR vaccine.
For the remaining 18 cases the
diagnosis of a gastrointestinal
disorder was not medically confirmed.
4. Discussion
The original remit of
the Working Party envisaged that it
might place its findings in the
context of all the available relevant
evidence but this report only
considers evidence derived from the
evaluation of parental and medical
questionnaires. A wide variety of
possible adverse effects after MMR or
MR vaccines had been reported through
solicitors. The remit of the Working
Party was to consider "any of
the reported effects" but the
Working Party, having reviewed the
full range of possible adverse
effects reported by parents of 531
children, decided to consider those
which had been most frequently
reported and had caused most concern
to parents i.e. autism/PDD and
gastrointestinal disorders,
particularly Crohn's disease.
The Working Party
recognised that there were marked
limitations in studying the group of
children that had been identified by
the solicitors involved. These
children were not a randomly selected
or population-based group and there
were no controls. They had been
reported to solicitors because of a
perceived association between the
administration of MMR or MR vaccine
and the occurrence of various
illnesses.
A sample of 95 cases
with reported autism were selected
from the original 180 reports for
detailed evaluation. For cases of
autism/PDD with gastrointestinal
disorders and Crohn's disease, all
the cases with follow-up information
available by 1/8/98 were included.
For cases of autism/PDD alone,
selection was made on the basis of
earlier receipt of follow-up
information. This may have led to
some sampling bias but it is,
however, likely that any bias
introduced by this process will be
for more plausibly linked cases to be
included.
For 85% of the children
reviewed whose parents reported
autism, the available medical
information supported a diagnosis of
PDD or autism. Autism/PDD was
considered possible in a further 12%
of cases. Developmental assessments
and clinical examinations were
normally provided. Results of
investigations such as hearing tests
and chromosome analyses were often
reported. The confirmed cases showed
no extraordinary characteristics and
there was no evidence to support a
novel syndrome associated with MMR
vaccine.
In assessing the
temporal significance of reports in
relation to vaccination, a period of
onset within six weeks was considered
of particular interest because this
is the maximum period in which viral
replication can be detected after
immunisation [1] and has been used
previously used for some of the
principal epidemiological studies of
MMR vaccine [2,3]. For 42% of cases
of autism, relevant symptoms were
first observed within six weeks
following vaccination according to
the parents. It is noteworthy that
parents were much more likely to
indicate that symptom onset was
within this interval than were
doctors. Only 5 of these children
were seen by doctors within six weeks
of immunisation. Where there was
disagreement between parents and
doctors we used the parental report
in our analysis although it would be
reasonable to assume that a dramatic
and acute deterioration in a child's
neurological, behavioural or
cognitive state would have prompted
medical consultation. Often the
doctor only provided the date of the
first consultation when symptoms were
first reported to him or her, rather
than the actual time of onset.
Because of the gradual way in which
autistic features evolve in a child,
it is not surprising that the time of
onset was not known for many
children.
Comparisons of the
characteristics of children whose
symptoms were reported to have had an
onset within or greater than 6 weeks
following immunisation revealed no
differences. For the assessment of
cases of autism/PDD there was some
interobserver variation, with
complete agreement on all four
categories for 24% of cases and all
but one category for 63%. Differences
were fairly evenly spread amongst the
four categories and are likely to
reflect the considerable difficulty
involved in interpreting the variable
and often limited information
reviewed.
In the children with
confirmed or possible autism there
was no evidence suggestive of a
preceding acute neurological event
such as encephalitis or
encephalopathy, either before or
after MMR vaccine. With regard to
previous medical history suggestive
of behavioural difficulties, there
was a marked contrast between the
frequency estimated from information
provided by parents (1%) and doctors
(39%).
There were 28 cases
(30%) recorded of speech or
behavioural difficulties in childhood
affecting first-degree relatives of
autistic children (with multiple
instances in some families). There
were also seven children with
abnormal obstetric histories and six
with a preceding possible viral
illness which had been considered
relevant either by the parent or the
reporting doctor.
When considering the
bowel symptoms associated with
autism, there were no consistent
clinical features or evidence of a
close temporal relationship with MMR
vaccine. Some of the findings
described on endoscopy such as
lymphonodular hyperplasia may be
normal features [6]. In particular,
there were no reports with adequate
evidence to support a diagnosis of
inflammatory bowel disease occurring
in association with autism. No
consistent recognised or novel
gastrointestinal syndrome could be
defined in these children with
autism/PDD and gastrointestinal
symptoms.
Crohn's disease was
usually well documented
diagnostically and was considered
confirmed or possible in 15 cases.
Five cases of Crohn's disease had an
onset of symptoms within six weeks
after vaccination according to the
parents. Four of the cases were
categorised as "YYNN" but
no differences in characteristics
between these children and the others
with Crohn's disease were identified.
The Working Party
carefully considered whether further
evaluation of cases of reported
autism/PDD and Crohn's disease not
yet reviewed in detail should be
undertaken. Taking into account the
high proportions of cases already
sampled with the diagnoses of
particular concern, the absence of
findings suggesting a causal
association and the limitations of
the evidence (in particular selection
bias leading to an expected temporal
association), the Working Party
considered that further detailed
evaluation is not warranted.
5. Conclusions
of the Working Party
Based on their review of
information described above, the
Working Party reached the following
conclusions :
5.1
The exercise undertaken in respect of
cases of children with adverse
effects attributed by parents to MMR
or MR vaccines and reported to a firm
of solicitors has yielded a
considerable volume of medical
information which was extremely
variable in quality and completeness.
5.2
The information evaluated has
important intrinsic limitations as
regards assessing whether the
vaccines are, or are not, causally
associated with the attributed
adverse effects. Notably these are
bias in the selection of cases for
which it would be impossible to
compensate, and the lack of any
control (unimmunised) group with
which the frequency and
characteristics of the attributed
illnesses could be compared. Also
there was frequent divergence between
parents and doctors regarding
specific details of the illnesses.
5.3 Detailed
evaluation of 92 cases with
autism/PDD and all 15 cases with
confirmed Crohn's disease revealed no
extraordinary features which
suggested a novel syndrome, nor did
it support causal associations with
MMR or MR vaccines. In particular, no
case of autism/PDD developed
following an acute unexpected and/or
unexplained neurological event in the
post-vaccination period.
5.4
For only eight cases of the 92 with
autism and four of the 15 with
Crohn's disease was satisfactory
evidence available to infer (i)
medical confirmation of the diagnosis
(ii) a close temporal association
between administration of the vaccine
and onset of the illness (iii) no
relevant prior history and (iv)
absence of an alternative cause. The
numbers of such cases identified was
therefore small. Furthermore these
factors alone are insufficient to
prove causation, particularly as the
onset of autism is frequently
recognised around the time MMR
vaccine is given. The pattern of the
illnesses reported for the 8 cases
with autism and 4 with Crohn's
disease cited above did not appear to
differ from that of other children
with these disorders. There was no
evidence to suggest that
administration of MMR or MR vaccine
was associated with particular
variants of pervasive developmental
disorder or inflammatory bowel
disease.
5.5
It was impossible to prove or refute
the suggested associations between
MMR vaccine and autism/PDD or
inflammatory bowel disease because of
the nature of the information, the
self-selection of cases and the lack
of comparators. Nevertheless, the
Working Party found that the
information available did not support
the suggested causal associations or
give cause for concern about the
safety of MMR or MR vaccines.
References
1. Mahy BWJ, Collier L
(eds). Topley and Wilson's
Microbiology and Microbial Infections
(9th Edition) Volume 1 : Virology ;
Arnold, London, 1998 ; 446-447.
2. Miller E, Goldacre M. Pugh S. et
al. Risk of aseptic meningitis after
measles, mumps and rubella vaccine in
UK children. Lancet 1993 ; 341 :
979-982.
3. Farrington P, Pugh S, Colville A
et al. A new method for active
surveillance of adverse events from
diphtheria/tetanus/pertussis and
measles/mumps/rubella vaccines.
Lancet 1995 ; 345 : 568-569.
4. Szatmari P, Jones MB, Zwaigenbaum
L, Maclean JE. Genetics of autism:
Overview and new directions. Journal
of Autism and Development Disorders
1998 ; 28: 351-368.
5. Rutter M, Bailey A, Siminoff E,
Pickles A. Genetic influences and
autism. In: Cohen DJ, Volkmar FR.
Handbook of autism and pervasive
developmental disorders (2nd edition)
Wiley, New York 1997 ; 370-387.
6. Bartram CI, Halligan S.
Radiological investigation of chronic
inflammatory bowel disease in
childhood. In: Walker-Smith JA,
MacDonald TT. Balliere's Clinical
Gastroenterology Volume 8 No. 1 ,
Balliere Tindall, London, 1994 ;
101-111.