Late last month 30 top scientists warned in an open letter: "The time has come to draw a line under the question of any association between the MMR vaccine and autism. The UK's children are in danger of serious illness or death if they are left unimmunised."
But Anna's death raises another question: whether these same health authorities are digging deeply enough into reports of serious illness, death -- and yes, autism -- following MMR vaccination.
The facts are still being determined in Anna's case, but John Duncan said his daughter had been exposed to chickenpox at a party just before she got the combined measles-mumps-rubella shot. She broke out with classic chickenpox just days after the shot and died a week later -- on Sunday, May 14 -- from an apparent seizure.
Duncan said that Anna had a runny nose when she got the MMR shot on April 26 and that his wife, Veronica, told the healthcare worker that Anna had been exposed to chickenpox. The response: No problem.
John Duncan couldn't disagree more, and neither could we. Our reporting suggests too many viruses -- from simultaneous live-virus vaccines, viral infections or some combination thereof -- may be a very big problem in very young children. The Age of Autism series Pox, which by happenstance ran at the same time as Anna's chickenpox exposure and illness, identified a cluster of autism cases in Olympia, Wash.
The common thread: The children came from families with problematic chickenpox histories and got closely timed MMR and chickenpox vaccines at a young age, in most cases their 12-month checkups. (Chickenpox vaccine is recommended in the United States starting at 12 months; it is not administered in Britain.)
If the observation were borne out in broader studies, it could mean combined live viruses in vaccines interact in a way that triggers autism in susceptible children.
That's not happening, according to public health authorities; they say a link between vaccines and autism has been ruled out.
Yet as far as we can tell, there's no research -- zilch, nada -- on whether inherited susceptibility to certain viruses could affect a child's capacity to handle the weakened, or attenuated, versions of those viruses in vaccines.
On top of that, we recently came across two old and overlooked but potentially disturbing studies. The first is titled "Viral Exposure and Autism," by Eva Y. Deykin of the Harvard School of Public Health and Brian MacMahon, in a 1979 (yes, 1979) issue of the American Journal of Epidemiology.
The authors looked at whether exposure to certain viruses, either in utero or in infancy, can be linked to autism. Here's what they said:
"The data indicates that total autistic symptomatology seems to be associated with prenatal viral experience with measles and mumps and with infancy illness or exposure to mumps and chickenpox." The study also reported a widely recognized connection to prenatal rubella (German measles) exposure.
The authors noted: "The proportion of cases that could be accounted for by any of these associations, even if causal, is small."
Still, a study a quarter-century ago saw a possible link between autism and viral exposures -- the same viruses that are at issue in the Olympia cases and in Anna Duncan's death.
The second study is from a 1973 (yes, 1973) issue of The Lancet -- a prestigious British medical journal that ought to carry buckets o' credibility with that country's medical authorities. It looked at SSPE, a delayed and often deadly brain infection that occurs in a small percentage of children who get measles.
The authors, led by Roger Detels of the Division of Epidemiology at the UCLA School of Public Health, wanted to identify risk factors for developing the subsequent brain infection. The study matched 38 children who had SSPE with 38 who did not and came up with several interesting observations. Here is the key:
"Among the 38 matched pairs, 6 patients had had chickenpox six months or less before measles. ... This sequence did not occur among controls."
This amounts to a "significant excess of chickenpox associated with measles in SSPE patients," the study concluded. "While this occurred in only six instances it is of note because of the relatively early age of clinical measles in patients versus controls, decreasing the likelihood of this sequence."
So, chickenpox followed by measles in young children might trigger brain damage and death? Troubling, indeed, in light of what happened to Anna Duncan.
Let's be clear -- neither study is big enough to be definitive, and neither is about vaccine viruses. But both point to neurological consequences when those same viruses infect children at the same early ages, in particular when they occur close together.
Interaction between viruses is in fact a well-known issue with vaccines, too. As we reported in our Pox series, a first-ever MMR-chickenpox vaccine from Merck & Co. called ProQuad has 10 times the standard dose of chickenpox virus in order to overcome a phenomenon called immune interference.
In 2004 a Merck scientist said the chickenpox component of ProQuad might be "causing a local immune suppression and an increase in measles virus replication," according to minutes from a meeting at the Centers for Disease Control and Prevention.
One of the Olympia children diagnosed with autism had been in a clinical trial of ProQuad, which was approved by the U.S. Food and Drug Administration last September.
The Olympia cases, Anna Duncan's death and the two 1970s studies belong to what I call a "canon of caution" about viruses, live-virus vaccines and brain damage -- and regrettably it is getting larger every day. A parallel case is also in the news in Britain: In January, an 18-month-old named George Fisher died in his sleep 10 days after the MMR, and an inquest has just been postponed to gather more evidence; the parents blame the MMR.
When scientists like the "U.K. 30" dismiss these anguished personal concerns, one wonders how thoroughly they have studied the canon themselves -- or if they even know it exists. (A lawyer told me that when concerns began in Britain in the late 1990s about the MMR and autism, he begged government health officials to come look at the disturbing parental accounts he had compiled. They declined, he said.)
You can adjust childhood immunization policy -- say, separating the MMR components and giving them a year apart as some critics suggest -- and still protect children from deadly epidemics. But you can't adjust the outcome of a fatal seizure.
Those in charge need to demonstrate they are as concerned about the health of every single child who gets a shot as they are about "herd immunity." That, ironically, is the only real way to restore Britons' flagging confidence in vaccines and head off the return of dangerous diseases.
And that would be a fitting legacy for Anna Duncan.