Measles, Mumps, Rubella Vaccine: Through a Glass, Darkly.

January 23, 2001

(Adverse Drug Reactions 2000,19(4) 1-19. Authors: Andrew J Wakefield FRCS,
Scott M Montgomery PhD.)

      Confidence in vaccine safety is essential for the success of
vaccination programmes.
      There is growing parental and professional concern about the safety of
the trivalent MMR vaccine.
      Suspected long-term adverse events from MMR include some forms of
regressive autism and inflammatory bowel disease (Crohn's disease and
ulcerative colitis).
      Some doctors and scientists suspect that the combination of the three
live viruses in one vaccine increases the risk of these adverse events. In
1999 we reported data - endorsed by the Medical Research Council, peer
reviewed and published - showing that concurrent exposure to natural (or
vaccine) measles and natural mumps infections is a significant risk factor
for inflammatory bowel disease.
      Vaccine manufacturers and departments of public health unanimously
assert the safety of the trivalent measles, mumps and rubella vaccine,
citing extensive safety testing. Moreover, the public is assured that there
is nothing in the safety studies to provide a basis for any anxiety over
possible delayed adverse effects.
      This important new article examines the safety trials of MMR and finds
no grounds for complacency. Instead the adverse autistic and
gastrointestinal side-effects now asserted by tens of thousands of parents
in the US, UK and elsewhere are foreshadowed in the results of medical
research and of the safety trials even before the vaccine was introduced.
      *  Pre-licensing trials of MMR revealed gastrointestinal events that
persisted to the end of the trial period in significant numbers of children
from developed countries. Despite this, the follow up period for subsequent
trials was reduced from 28 days to 21 days.
      Furthermore, in subsequent trials the data from contrasting
populations (developed and developing countries) was aggregated, masking the
potential significance of gastrointestinal side effects in children from
developed countries.
      *  The decision to combine the three vaccines in one (undoubtedly
atypical) exposure was taken without specific consideration of the known
associations between concurrent exposures to common childhood infections and
later consequences. By 1979, atypical patterns of exposure to measles,
mumps, rubella and chickenpox were identified as possible risks for autism.
      *  In addition to the above, the very first pilot studies of MMR
revealed that the component viruses of MMR could interfere with one another.
      Despite evidence of potential dose and strain-dependent interactions
affecting immune responses to the vaccine viruses and therefore possible
adverse events, the question was never investigated further.
      *  The import of the article is enhanced by the publication of the
comments of four reviewers. Taken as a body, the reviews are highly
supportive of this new paper. Two reviewers, in particular, concur that, on
the evidence presented by Wakefield and Montgomery, the safety trials of MMR
were at best weak and at worst inadequate.
      Clearly in the context of MMR one plus one plus one never did equal
      In summary, the authors suggest that until such time as the question
of MMR safety is resolved, the public must, at the very least, be offered a
      As the last Minister for Health, the Hon Frank Dobson said, in another
medical context, "If there is even a hypothetical risk (of harm) and a safer
alternative exists, we should use it".
      In the current circumstances, given the significant index of suspicion
that now exists in relation to the safety of MMR, regressive autism and
inflammatory bowel disease (Crohn's disease and ulcerative colitis), then it
is surely right to allow parents to use the monovalent vaccines where they
are, rightly or wrongly, concerned.
      Perhaps the strongest endorsement of the value of this paper comes
from one of the reviewers, Dr A Peter Fletcher, MB BS PhD, a former
Principal Medical Officer in the Medicines Division, now MCA, of the
Department of Health who served as Medical Assessor to the Committee on
Safety of Medicines.
      "With all the benefits of hindsight, what may now be said about the
decision to grant a Product Licence (as they were then called) to MMR 10 or
so years ago? Evidence on quality and efficacy was probably adequate so a
decision had to be made on grounds of safety.
      Being extremely generous, evidence on safety was very thin. Being
realistic there were too few patients followed up for insufficient time.
Three weeks is not enough, even for RCT's, neither is 4 weeks. By 1988-89 we
knew from experience with pertussis vaccination that longer duration was
essential- how much longer it is difficult to say but as long as humanly
      We also knew that numbers, big numbers were equally necessary.
Additionally we knew that observational cohort studies could be conducted on
10,000 or more patients for up to 18 months. Primary care computerised
databases (GPRD for example) were already up and running which would permit
prospective record surveillance on several million patients. There was
insufficient information on the immunological effects of a trivalent vaccine
compared to monovalent vaccines.
      Was there detectable immunosuppression with trivalent vaccine versus
monovalent? From known clinical experience with measles mumps and rubella
infections we could make an estimate of the incidence of serious disease
outcomes which would be prevented by effective vaccination. From these
figures we could make an informed guess of ADR levels that could be
      Did the available evidence on the trivalent vaccine support the belief
that benefit would outweigh risk? On the basis that effective monovalent
vaccines were available the CSM could be confident that delay in granting a
licence would not result in a catastrophic epidemic of measles, mumps and
      Caution should have ruled the day, answers to some important questions
should have been demanded and strong encouragement should have been given to
conduct a 12-month observational study on 10-15,000 patients and a
prospective monitoring programme set up with a computerised primary care
database. The granting of a Product Licence was premature."

      One of the authors of the paper, Andy Wakefield FRCS, makes the
following additional comments.
      "This paper is a summary of a detailed analysis of safety trials for
measles containing vaccines that I began in 1994."
       "The Department of Health (in the United Kingdom) were in receipt of
an earlier draft version of this paper, which was sent to them some months
ago in order to permit adequate time to prepare sufficient stocks of
monovalent vaccine pending publication of the paper."
      "I am a firm believer in the protection of children against serious
infectious diseases and their consequences. This paper advocates vaccination
against measles, but the issues of vaccine safety are vitally important and
must be of paramount concern".