#158-Brian Deer: A tale of scientific fraud—Andrew Wakefield & the belief that vaccines cause autism

KEYPOINTS

• 01

  The MMR–autism scare grew from one researcher's fixation on measles.

  Deer, the investigative journalist who exposed Andrew Wakefield, traces the story to the late 1980s. Wakefield, trained as a surgeon, observed early whole-bowel transplant attempts at Toronto General Hospital, then moved into research. Influenced by the AIDS era's search for infectious agents, he theorized that Crohn's disease was a disease of the gut's blood supply caused by an infection, and settled on measles virus.

• 02

  Wakefield clung to his hypothesis when others couldn't replicate it.

  In 1993 he published in the Journal of Medical Virology claiming measles in Crohn's tissue using immunohistochemistry. About two years later a Japanese group using PCR, the more specific emerging gold standard, couldn't find it. Wakefield argued PCR missed it, yet later switched to PCR himself when it suited the MMR claims, which Deer sees as evidence the conclusion came first.

• 03

  The research was set up to serve a lawsuit.

  Lawyer Richard Barr, building a case against MMR manufacturers, hired Wakefield in early 1996. In June 1996 they applied to the Legal Aid Board, a government fund that paid for people suing drug companies, and the grant application promised PCR findings before any child in the eventual paper was investigated. Deer says none of this was known until his Sunday Times investigation.

• 04

  The 1998 Lancet paper was a 12-child case series—and the format wasn't the problem.

  Published in February 1998 in one of medicine's top journals, it reported that parents of eight of 12 children (two were brothers) said symptoms began within 14 days of MMR. Attia and Deer note case series are legitimate: Crohn's disease was characterized from 14 patients, autism from 11 children at Johns Hopkins, and AIDS from five men in Los Angeles in 1981. The issue was what was misrepresented.

• 05

  The children were pre-selected through the litigation network.

  None of the 12 lived in London; the nearest was about 60 miles away and some were 280 miles. Many came via a campaign group whose founder, Jackie Fletcher, had a child who became Wakefield's 'sentinel case'. Autism had only recently become a formal diagnostic category in the early 1990s.

• 06

  Not one child's case was accurately reported.

  When Deer obtained the medical records, some children's developmental issues predated vaccination, others began months after, not within 14 days, and some reported diagnoses didn't match. The sentinel case's mother admitted in court papers that her child showed no issues for nine months after vaccination.

• 07

  PCR is powerful, but past about 35 cycles a 'positive' is unreliable.

  Each PCR cycle doubles the target, so 35 cycles means up to about 2^35 copies. Because the method is so sensitive, contamination produces signals at high cycle counts; a negative is almost certainly negative, while a late positive is uncertain. Attia notes reporting cycle thresholds matters: positive at 20 versus 30 cycles reflects a very different starting amount.

• 08

  At the epicenter of the fraud were manipulated machine outputs.

  Biopsies went to a lab in Dublin run by John O'Leary, using an ABI Prism 7700, certified as a research tool and not for clinical diagnosis. Its 96-well trays held samples, positive measles controls, distilled water and empty wells. After drug companies' lawyers obtained the data through the London and Irish High Courts, an expert, Stephen Bustin, found wells the machine had read as positive reported as empty and empty wells reported positive—manual changes that made results fit Wakefield's thesis.

• 09

  Wakefield's own scientist found nothing.

  A second set of samples went to Wakefield's researcher Nick Chadwick, who froze them in liquid nitrogen; testing found no measles virus. Chadwick later left science, Deer believes partly from disillusionment.

• 10

  The consequences came late but were decisive.

  A General Medical Council hearing ran for 217 days, longer than O.J. Simpson's trial, costing about £10 million. Wakefield lost his license and the Lancet retracted the paper in 2010. The litigation consumed roughly $100 million in today's money and never reached trial.

• 11

  Autism's diagnostic history made it easy to misattribute.

  Deer stresses that autism is a construct—a checklist diagnosis clustering signs and symptoms—rather than a disease with a single lesion like Crohn's. Formal categories from the American Medical Association and World Health Organization arrived only in the early 1990s, so diagnoses were rising just as Wakefield's network went looking for vaccine-injured children.