Saturday, May 29, 2010

it's not about the science

It's a rare day when Andrew Wakefield gets struck off and Nick Griffin is forced out of the BNP leadership. But this will not do:
Today's verdict ‑ the striking-off of Wakefield and Prof John Walker-Smith, who was in charge of the department of paediatric gastroenterology at the Royal Free hospital in London, where the research took place and the acquittal of the-then junior consultant Simon Murch, who had doubts about the project ‑ was about ethics and honesty, not science.

It was about the attitude of the doctors to their child patients. While Walker-Smith made errors of judgment and ultimately paid the price because of his position of responsibility, Wakefield emerged from the hearings as a tarnished character, branded dishonest. He was found to have subjugated the needs of vulnerable children to his desire to prove a theory....

But Wakefield's disgrace will not stop him arguing over the science, flawed as experts say his arguments are. Based in the US, where he will still be able to work as a scientist ‑ if not a doctor ‑ and with a considerable following among the desperate parents of autistic children who get too few answers about the distressing illness, he will continue to portray himself as the victim of the British medical establishment. It is not all over yet.

As Anthony Cox says,
However, while Wakefield has gone, the media environment that allowed him to drive down vaccination rates continues to exist.

Bizarrely, the papers - not even the Grauniad - still seem unable to even mention the fact that Wakefield's results were worthless. For that, you've got to go to some random blog - like Holfordwatch, devoted to harassing the tiresome media nutritionist of the same name. Fortunately, Mr. Holford decided to get involved in the Wakefield affair, and so some of Holfordwatch's light was shed on Wakefield. As they rightly say:

But Holford is concerned about the measles antibodies in the cerebro-spinal fluid and the gut tissue of the children Wakefield examined; Holford believes that Wakefield found evidence of this. Whatever significant problems there may be with the rest of Wakefield’s work, surely this is the crux of the matter?

Indeed; if there was evidence of measles virus, then there is at least the faintest possibility that he might have been on to something. But there was no such evidence, or rather, the evidence he presented of them was spurious. Dr. Nick Chadwick, Wakefield's student in the late 90s, actually carried out most of the tests for measles RNA - and he repeatedly got either negative results, or else results that turned out to be false positives on closer scrutiny.

Chadwick and Wakefield were using the polymerase chain reaction (PCR) test to look for measles viruses. It works by using the enzymes that copy DNA sequences in nature, causing base molecules to polymerise - link together in strings - in the same order as they are in the original sequence. Many millions of years of evolution went into this process, and it is very efficient indeed at copying the nucleic acids. There is the rub. If there is any DNA at all anywhere near the sample, PCR will copy it, and once copied, it will copy it again, and again, until the concentration is high enough for other methods to be used.

The danger of high sensitivity is the creation of false positive results. It is absolutely essential to success that experimenters with PCR take truly obsessive precautions against cross-contamination. In order to check the validity of the positive results, Chadwick sequenced the RNA (measles viruses don't have DNA - technically, what he was doing was RT-PCR, using a reverse transcriptase to copy the measles RNA) from the positive results and compared it with that in the MMR vaccine. If the viruses were those from the vaccine, the sequences would be identical; if not, the result was invalid, as whereever they came from, it obviously wasn't the vaccine. He also compared the sequences with samples of measles collected in the wild.

All the nine hits achieved with PCR turned out to be false positives.

Wakefield was also sending samples elsewhere for analysis; Chadwick found that, again, the results were false positives, and that the lab in question was producing an unusually high false positive rate, evidence of possible contamination. Also, some of the samples sent away were duplicates, included as a check - if measles was present in one duplicate, it would also be present in the other, so both of them would be hits. If only one of the duplicates was a hit, this would suggest that something was wrong.

He formed a theory that the measles antibodies they were using to detect the virus were suspect, and might be reacting to one of the gut bacteria rather than measles, which if true would have required tearing up all the results and starting again. The sequences did match an artificially bred strain of measles, unknown outside the lab environment, which strongly suggested that some of the material or equipment in use was contaminated.

But none of this bothered Wakefield, who pressed on; Chadwick eventually refused to be cited as a co-author of the paper Wakefield sent to the Lancet. The whole sorry story is on the public record - Chadwick testified in a lawsuit in the United States in 2007, and his testimony is here. Elsewhere in the case, there's also an interesting technical discussion (go to page 145) about PCR and false positive results.

Can anyone guess why the British press doesn't want to discuss this part of the story?

No comments:

kostenloser Counter