The MMR scare: from foolishness to fraud?

‘What may begin as honest error, however, has a way of evolving through almost imperceptible steps from self-delusion to fraud. The line between foolishness and fraud is thin.’ (Robert Park, Voodoo Science, 2000.)

Brian Deer’s allegations in The Sunday Times, on 8 February, that Andrew Wakefield fixed data in the 1998 Lancet study that launched the scare linking the MMR vaccine to autism suggest that the former Royal Free researcher is facing the wrong charges in the ‘professional misconduct’ proceedings currently underway at the UK General Medical Council (GMC).

From the very start of this scandal, I have worked to debunk the myth of a link between MMR and autism. However, when it emerged nearly three years ago that Dr Wakefield was to be summoned to appear before the GMC on charges of professional misconduct, I argued that this was ill-advised (see Stop witch-hunting Wakefield). Though I accepted that he had a case to answer in relation to the ethical propriety of conducting invasive investigations (ileo-colonoscopy, lumbar puncture) on autistic children, and in relation to undisclosed legal aid funding, these did not appear to me to justify such punitive proceedings. The GMC hearings, which opened in July 2007 and are set to continue into the summer, have confirmed that Wakefield has considerable room to wriggle around these charges. These hearings have had the additional disadvantage of implicating the former Royal Free paediatricians Simon Murch and John Walker Smith, decent clinicians who have paid a heavy price for loyalty to Dr Wakefield. Even if Dr Wakefield is struck off the medical register by the GMC, he will still be able to claim that the science advanced in his Lancet paper is sound.

By contrast, Brian Deer’s allegations go the heart of the scientific basis for the claim for a link between MMR and autism, mediated by the hypothesised inflammatory bowel condition subsequently dubbed ‘autistic enterocolitis’ by Dr Wakefield.

One of the key claims of the Lancet paper was of a close temporal association between the children receiving their MMR jab and the onset of autistic symptoms. Of the 12 cases, parents of eight linked the onset of ‘behavioural problems’ to MMR: ‘In these eight children the average interval from exposure to first behavioural symptoms was 6.3 days (range 1-14).’ (1) According to Deer’s account of the records presented to the GMC, in some of these children autistic features were present before they had MMR and in others such features appeared after a much longer interval than a few days after MMR. Deer found significant differences between the children’s medical records and the history presented in the Lancet paper.

Another key claim was that Wakefield and colleagues had found a distinctive pattern of gut inflammation – ‘colitis and ileal-lymphoid-nodular hyperplasia’. The Lancet paper emphasised the ‘uniformity’ of the intestinal pathological changes and that the gastrointestinal findings were ‘consistent’. This interpretation was disputed by Susan Davies, the Royal Free pathologist who examined gut biopsy specimens from the cases in the Lancet study. She told the GMC that the biopsy results were not uniform, but varied – and in most cases the microscopic findings were unexceptional. She expressed particular concern at the use of the term ‘colitis’ to describe what she regarded as ‘relatively minor’ abnormalities. Indeed, in seven of the 12 cases, laboratory examination had first reported no significant abnormality. It was only on subsequent review that these results were re-interpreted as abnormal. (It is not clear why it has taken more than a decade for such serious allegations, in relation to a study which – as Deer documents – has had such a damaging effect on families affected by autism as well as on child health more broadly, to emerge into the public realm.)

Another important theme in the Lancet paper was that the 12 cases included had been ‘consecutively referred’ to the Royal Free. It has long been clear that most of these children came to be referred to the Royal Free because of their involvement in the campaign against MMR led by the solicitor Richard Barr and anti-MMR parent groups. Not one of the children came from the Royal Free’s North London catchment area – indeed none came from anywhere in Greater London. The cases came from all over the country – one came from the USA. Deer’s investigations confirm that 11 of the 12 cases were included in the anti-MMR litigation (that collapsed in the absence of supportive evidence in 2003). These cases were thus highly selected and could not be regarded as a random sample of children with autism or even children with autism and bowel problems.

A fourth problem in the Lancet paper is touched on, but not fully developed, in Deer’s most recent article. In a much-quoted sentence in the concluding section, the authors concede that ‘we did not prove an association between MMR vaccine and the syndrome described [autistic enterocolitis]’. The following sentence states that ‘virological studies are underway that may help to resolve this issue’. Deer reports Dr Wakefield’s comment at the press conference to launch his Lancet paper in February 1998 that ‘we would not have presented this paper to the Lancet had we not undertaken extensive virological studies already’.

In fact, as Deer reported in his 2004 Channel 4 documentary, Dr Wakefield had already received results of investigations of gut biopsy specimens for measles from the 12 Lancet cases carried out by Nicholas Chadwick, then a junior member of his team. But because these results had all been negative, Dr Wakefield refused to include them in the Lancet report. (Chadwick insisted on his name being withdrawn from the paper. He provided evidence on this matter to the Omnibus Autism proceedings in the USA in July 2007.) Deer reports how the father of one of the Lancet cases took biopsy specimens for measles studies at independent labs in the UK and the USA – all were negative.

Deer’s claims that Dr Wakefield manipulated and effectively falsified key data in the Lancet study are not the only such allegations that have been made against his research. Three other Wakefield papers have also been subjected to serious criticisms along similar lines.

In a study published in the American Journal of Gastroenterology in 2000, Dr Wakefield and his Royal Free colleagues presented a series of 60 ‘consecutive’ cases of ‘autistic enterocolitis’ (including the original ‘Lancet 12’), claiming ‘an endoscopically and histologically consistent pattern of ileo-colonic pathology’ (2). This paper was subjected to a withering critique by Professors Tom MacDonald and Paola Domizio, from the Barts and the London medical school (3). These authors observed that Wakefield had failed to demonstrate ‘enteritis’ as he claimed and that what he labelled ‘colitis’ was the sort of mild inflammation commonly associated with constipation, which was a feature of many of these children. They also raised serious questions about the manipulation and misrepresentation of data.

Having shelved Chadwick’s negative measles studies, Wakefield sent gut biopsy specimens to the laboratory of the Dublin pathologist John O’Leary, with whom he collaborated on two papers published in 2002. The first (lead author Uhlmann) claimed to have identified measles virus in a high proportion of these specimens; the second (lead author Martin) claimed to have identified ‘vaccine-strain’ measles virus, thus directly implicating MMR. Though both these studies received a sceptical response from experts, they had a major impact in the media, and encouraged many parents to pursue litigation claims (like Wakefield, O’Leary received substantial remuneration from the Legal Aid Board).

Another witness to the Omnibus Autism proceedings in Washington in summer 2007 – Professor Stephen Bustin, also from Barts and the London medical school – presented more than 100 pages of evidence based on his investigation of the O’Leary lab (4). Bustin’s conclusions were categorical: ‘The assay used was not specific for measles and it was not properly carried out.’ The positive results were positive for DNA – confirming contamination, because ‘if it’s DNA it can’t be measles’ (measles is an RNA virus). For Bustin it was ‘a scientific certainty’ that the O’Leary lab had failed reliably to identify measles virus RNA in any child. Bustin’s devastating testimony effectively destroyed the only piece of positive evidence that has been produced in support of the MMR-autism thesis. The multiple errors of these studies did not occur randomly, but all tended towards producing the positive results for measles virus that the investigators wanted to produce. That raises the question of something beyond junk science.

Following Brian Deer’s 2004 revelations about Dr Wakefield’s conflicts of interest arising from undisclosed legal aid funding, 10 of his Lancet co-authors retracted the suggestion of a link between MMR and autism (while upholding the paper’s claim to have identified a distinctive form of bowel inflammation in autistic children). It is now clear that, given the selection bias confirmed by Deer – quite apart from his other allegations – it is not possible to make such a claim on the basis of the Lancet cases. Surely it is now time for the authors to withdraw this paper in its entirety? Perhaps the editor of the Lancet – together with those of the other journals involved – could submit Deer’s allegations to some sort of tribunal, perhaps arranged by the Medical Research Council. For 10 years the world of science has witnessed Dr Wakefield’s foolishness; now it has to ask: has he crossed the line into fraud?

Once again families affected by autism owe a debt to Brian Deer, who has sat through the GMC hearings despite enduring a torrent of scurrilous abuse by anti-MMR campaigners. Almost alone among British journalists, he has been prepared to ask difficult questions and as a result he has got to the heart of the MMR controversy.

Dr Michael Fitzpatrick is author of the new book Defeating Autism: A Damaging Delusion, published by Routledge. (Buy this book from Amazon(UK) or Amazon(USA).)

(1) Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children, AJ Wakefield et al, Lancet, 351, February, 1998

(2) Enterocolitis in children with developmental disorders, AJ Wakefield et al American Journal of Gastroenterology, 95 (9), February, 2000

(3) Autistic enterocolitis: is it a histopathological entity?, MacDonald and Domizio, Histopathology, 50 (3), January, 2007

(4) The MMR- autism theory – there’s nothing in it, by Dr Michael Fitzpatrick, 4 July 2007