Some monkey business in autism research, 2009 edition

ResearchBlogging.orgAs you may have noticed, I’ve fallen into a groove (or, depending on your point of view, a rut) writing about anti-vaccine lunacy. The reason is simple. While I was busy going nuts over Bill Maher’s receiving the Richard Dawkins Award, the anti-vaccine movement has been busy, and there are some things I need to address that had backed up while I was distracted. There’s one more thing I need to address before I move on to other topics.

Over the last couple of months, I’ve noticed something about the anti-vaccine movement. Specifically, I’ve noticed that the mavens of pseudoscience that make up the movement seem to have turned their sights with a vengeance on the Hepatitis B vaccine. The reason for this, I believe, is fairly obvious. The fact that the Hep B vaccine is administered shortly after birth seems somehow to enrage the anti-vaccine movement more than just about any other vaccine. Moreover, given that, aside from maternal-child transmission when the mother is infected, hepatitis B is usually only contracted through either bloodborne contact (the sharing of needles, the administration of contaminated blood) or sexual activity, it’s very easy for anti-vaccinationists to make a superficially plausible argument that it’s not a necessary vaccine, even though there are reasonable rationales for giving it to infants. The image of sticking a needle into a newborn infant trumps that, though, at least for the anti-vaccine movement. Indeed, just the other day, the anti-vaccine crank groups the National Vaccine Information Center (NVIC), Talk About Curing Autism (TACA), and the anti-vaccine crank blog Age of Autism posted a call for the elimination of hepatitis B vaccination for newborns:

Washington, DC – National Vaccine Information Center and Talk About Curing Autism are calling on President Obama to order the immediate suspension of the Centers for Disease Control and Prevention recommendation of the birth dose of the Hepatitis B vaccine after two recent studies linking the Hepatitis B vaccine to functional brain damage in U.S. male newborns and infant primates.  In a related development today, the United States Department of Health and Human Services, including the Health Resources and Services Administration and Centers for Disease Control and Prevention,  announced that 1 in every 91 children are now diagnosed with an autism spectrum disorder as reported in the November 2009 issue of Pediatrics. Previous data released by the CDC indicated a prevalence of 1 in every 150 children affected by the disorder.


Note how AoA so cleverly interposed the latest information about autism prevalence with its call to eliminate the birth dose of the hepatitis B vaccine. Very clever. By doing so, it linked the two in readers’ minds, as if one had something to do with the other. Meanwhile, David Kirby is up to his usual nonsense. At the heart of this are two studies, one a restrospective study in humans and the other a study in monkeys, both of which the anti-vaccine movement is promoting as slam dunk evidence that the hepatitis B vaccine is causing all sorts of horrific problems. Taking both of them on in one post is too much, even for my logorrheic tendencies. So I’ll deal first with the monkey study and then, either later this week or sometime next week, deal with the human study.

The reason I start with the monkey study is because, so confident are the anti-vaccinationists of the study that they’ve placed the accepted manuscript on the Thoughtful House website (PDF).The study has also been posted on the anti-vaccine blog Age of Autism (PDF). That means that you can read it for yourself. (Thanks, Thoughtful House and Age of Autism!) The study is entitled Delayed Acquisition of Neonatal Reflexes in newborn Primates receiving A Thimerosal-containing HepatitiS B Vaccine: influence of gestational age and Birth weight, and it’s by a cast of characters that we’ve met before, including the pseudoscientist who launched thousands of cases of MMR in the U.K. through his shoddy science, being in the pocket of trial lawyers, and possibly even scientific fraud. Then there’s also Laura Hewitson. We’ve seen her before as the author of a couple of abstracts that I dissected in gory detail last year. Suffice it to say that, not only was the science shoddy, but massive conflicts of interest were not disclosed. The only difference this time around is that the conflicts of interest were disclosed. No doubt this was in response to all the criticism last year.

In fact, let’s look at the conflicts of interest first:

Prior to 2005, CS and AJW acted as paid experts in MMR-related litigation on behalf of the plaintiff. LH has a child who is a petitioner in the National Vaccine Injury Compensation Program. For this reason, LH was not involved in any data collection or statistical analyses to preclude the possibility of a perceived conflict of interest.

It’s hard not to retort that LH (Laura Hewitson) is the frikkin’ first author! She designed and coordinated the study! She’s also the corresponding author, which presumably means that she wrote the manuscript, or at least the bulk of it. There is not just the appearance of a conflict of interest; there is a conflict of interest, a massive conflict of interest. The same is the case with Andrew Wakefield. In fact, if I were to use the same criteria that Age of Autism did when it automatically labeled any study with any pharmaceutical company underwriting whatsoever as hopelessly biased in its Fourteen Studies website, I could stop right here and dismiss this study as so hopelessly the result of a conflict of interest that I don’t even need to analyze it.

Fortunately, unlike anti-vaccine cranks, that’s not how I roll. I did, however, have a hearty laugh at AoA’s attempt to justify the blatant conflict of interest thusly:

One likely tactic of critics of the study will include attempts to nullify the evidence based on the alleged bias of those involved. For one, the study is privately funded and acknowledges some well known autism advocates as financial contributors. These include the Johnson family (Jane Johnson is co-author of Changing the Course of Autism, a member of the Board of Directors of Thoughtful House and Director of Defeat Autism Now!), SafeMinds, the Autism Research Institute and Elizabeth Birt. Although all of these groups make clear their research interest is vaccine safety, they are frequently attacked for being “anti-vaccine”, an epithet that will almost certainly be hurled again here.

If the shoe fits…

However, even more amusing was this:

The most aggressive attacks, however, will likely be reserved for the study authors. The basis of these attacks is best anticipated by the following conflict of interest disclosure in the published paper.

After which Blaxill quotes the conflict of interest statement from the paper that I reproduced above. Isn’t the hypocrisy breathtaking? In Fourteen Studies, AoA and Generation Rescue slimed every investigator who received a penny of money from a pharmaceutical company or any governmental public health agency. In fact, Generation Rescu defined conflicts of interest this way:

We considered a scientist employed by a vaccine maker or a study sponsored by a vaccine maker to have the highest degree of conflict, with a public health organization (like the CDC) to be the second-worst.

Based on some of the things that Generation Rescue said about those “Fourteen Studies,” it also appeared to define the mere fact of being funded through grants from the CDC, NIH, American Academy of Pediatrics, or even the Canadian Institutes for Health Research as being a hopeless conflict of interest. Scientists who have had NIH grants (such as myself) or grants from any of these other organizations know just how ridiculous considering that particular funding source to be a horrific conflict of interest is. The bottom line is obvious. It’s a “conflict of interest” only if Generation Rescue says it’s a “conflict of interest.” Words apparently mean exactly what Generation Rescue says they mean, no more, no less. Another way of putting it is that it’s only a conflict of interest if the funding source happens to be an organization Generation Rescue hates. Accepting funding from a pharmaceutical company? It’s a conflict of interest. No doubt about it. Such COIs need to be reported, and–guess what?–they usually are. But, to Mark Blaxill and the Age of Autism, being one of the complainants in the Autism Omnibus and being funded by organizations whose purpose is to demonstrate that vaccines somehow cause autism and other neurodevelopmental problems? Well, that’s not a conflict of interest to anti-vaccinationists because they see themselves on the side of angels and think that they could never, ever have their objectivity affected by the funding source, having an autistic child, or being part of a legal action for “vaccine injury,” which, by the way, might be helped by scientific evidence showing that vaccines can cause autism or neurodevelopmental disorders. One notes that one of the things the previous monkey study published as an abstract by Hewitson and Wakefield got dinged for in the blogosphere was that no conflicts of interest were reported. Poor Mark seems really peeved at the criticism Hewitson justly received for that little ethical lapse last year.

On to the study.The first thing I always look at whenever reading any study is a simple question: What is the hypothesis being tested? This is the closest I could find:

Here we examine, in a prospective, controlled, observer-blinded study, the development of neonatal reflexes in infant rhesus macaques after a single dose of Th-containing HB vaccine given within 24 hours of birth, following the US childhood immunization schedule (1991-1999). The rhesus macaque is used in preclinical vaccine neurotoxicity testing and displays complex early neurobehavioral and developmental processes that are well characterized (reviewed by [7]).

Of course, to anyone who’s been involved in dealing with the anti-vaccine movement, one thing that’s very clear is that the subtext behind this is the unsinkable rubber duck of a belief among the anti-vaccine movement that, somehow, someway, either vaccines or mercury in vaccines causes autism. An inconvenient fact is that there has been no thimerosal in early childhood vaccines other than the flu vaccine since late 2001, but that doesn’t stop the anti-vaccine movement. I suspect that the reviewers of this article were probably blissfully ignorant of this context and concentrated solely on the methodology. Had they known, no doubt they would have asked some uncomfortable questions in their reviews. Of course, they would have no way of knowing that this study is in fact more of a propaganda tool than anything else. One thing that needs to be emphasized is that there really is no good primate model of autism, at least not that I’m aware of. That’s why Wakefield et al resorted to looking at infant reflexes.

Another question that needs to be asked. Why did the investigators look at thimerosal-containing hepatitis B vaccination? There’s no thimerosal in the hepatitis B vaccine anymore and hasn’t been since 2001. In fact, if you read the methods section of the paper, you’ll see that Hewitson et al added thimerosal to Recombivax HB (Merck) in order to recreate that thimerosal feeling from the 1990s. Why would they do that? Especially since the authors state in the conclusion that the study design “was not able to determine whether it was the vaccine per se, the exposure to thimerosal, or a combination of both, that caused these effects.” More on near the end of this piece.

Before I get to the effects observed, let’s just consider something else. When I read this study, there was something that set my skeptical antennae twitching fiercely. Remember the abstracts I discussed last year? Let’s take a trip down memory lane and read what I wrote back then:

What first leaps to mind in looking at the study is that there are 13 monkeys in the “vaccine” group and only three in the control group. No explanation is given for why there are such unequal numbers. Similarly, there is no mention of how the monkeys were assigned to one group or the other (randomization, anyone?), whether the experimenters were blinded to experimental group and which shots were vaccine or placebo, whether the monkeys were weight- and age-matched, or any of a number of other controls that careful researchers would do. Right off the bat, from the small numbers (particularly with only three monkeys in the control group), I can say that the study almost certainly doesn’t have the statistical power to find much of anything with confidence.

Now, let’s look at how many monkeys in this study: Thirteen receiving the hepatitis B vaccine plus added thimerosal. Well, now, doesn’t that seem rather…coincidental? There were also three animals receiving no injection and four receiving a saline placebo. Sound familiar? Well, there were three controls receiving no injection and four receiving saline placebo. Why do I bring this up? Remember, the abstract from last year described monkeys as undergoing the “entire vaccination schedule.” Remember how it was described that the vaccinated monkeys “exhibited autism-like symptoms”:

The first research project to examine effects of the total vaccine load received by children in the 1990s has found autism-like signs and symptoms in infant monkeys vaccinated the same way. The study’s principal investigator, Laura Hewitson from the University of Pittsburgh, reports developmental delays, behavior problems and brain changes in macaque monkeys that mimic “certain neurological abnormalities of autism.”

The findings are being reported Friday and Saturday at a major international autism conference in London.

Although couched in scientific language, Hewitson’s findings are explosive. They suggest, for the first time, that our closest animal cousins develop characteristics of autism when subjected to the same immunizations – such as the MMR shot — and vaccine formulations – such as the mercury preservative thimerosal — that American children received when autism diagnoses exploded in the 1990s.

Given the similarity of the description of the study described in the manuscript, in which hepatitis B vaccine was even spiked with thimerosal in order to mimick the vaccine schedule of the 1990s (given that hepatitis B vaccine no longer containes thimerosal) and the previously reported abstract, in which the entire vaccine schedule of the 1990s was supposedly mimicked, it does make me wonder. Could it be that the results being reported derived from observations made on the same monkeys? In other words, could it be that the investigators gave the monkeys the hepatitis B vaccine after birth, tested their various reflexes early in their lives, and then continued with their “simulated” vaccination schedule in order to produce the rest of the observations reported last year? Inquiring minds want to know! After all, the current study only goes out for 14 days after birth.

Indeed, one wonders if, stung by the criticisms of inadequate controls, the investigators added additional controls and kept the same group of 13 monkeys as the “vaccinated group.” Maybe they didn’t, but the similarity between the numbers of monkeys used in the abstract last year and the numbers of monkeys used in this study sure do raise an eyebrow. This is especially true given that in several of the analyses, the investigators pool the four monkeys receiving the saline control and the three receiving no injection for purposes of calculating means. Could it be that the investigators simply added a few monkeys after the experiment hda been started (or even after the original 16 monkeys had already undergone the entire “vaccination schedule”? Again, inquiring minds want to know! Could it be that, in order to beef up their apparent statistical power to detect differences in these various reflexes, some additional monkeys had to be added? Or was this done in response to reviewers’ concerns. If that’s the case, then when were these additional monkeys studied? How long after the original group. Mark Blaxill brags that the person who measured the monkeys’ reflexes was trained by an expert until her results had a high concordance with those of experts, but if there were a several month delay between when she measured the first group of monkeys and then the additional controls, it’s not too hard to imagine that she got better and thus more able to detect subtle differences in the reflexes. Again, the similarities between the designs of the studies described in the IMFAR abstracts last year and this study sure make me wonder if perhaps Hewitson and Wakefield are perhaps “minimizing” the use of animals. Of course, normally using as few monkeys as possible would normally be a good thing, but if that’s what they’re doing, why not report the entire study? Are they planning on having it come out in dribs and drabs, in other words posting several papers of what we call the “MPU” or “minimal publishable unit”?

In any case, in this particular MPU, what did Hewitson and Wakefield find? Not much, actually, the triumphant crowing of Mark Blaxill at AoA notwithstanding. Basically, Hewitson and Wakefield reported that three of thirteen reflexes were delayed, specifically the root reflex was delayed by one day; the suck reflex by nearly two days, and the snout reflex, also by nearly two days. There also appeared to be a confounding factor in that monkeys with lower gestational age (GA). For example, the authors state:

In general,as GA increased animals reached criterion earlier whereas animals of lower GA were relatively delayed. This effect was only significant when exposure was taken into account.

I really have to wonder whether in a larger group of unvaccinated monkeys, the correlation with these reflexes with gestational age would reach significance without taking the hepatitis B vaccine into account. Given such small numbers, I always wonder about any sort of multivariate analysis. Also, in this paper, these reflexes are ranked from 0 (absent) to 3 (the highest possible score). The time to criteria was defined as the time to reach the highest possible score. Again, given the small numbers and the correlation between gestational age and reaching these milestones, I really have to wonder if the results in this study, although statistically significant, are really behaviorally or biologically significant. If there’s one thing my mentors always taught me, it’s that statistically significant doesn’t necessarily mean significant.

Personally, I’m not all that impressed. One reason is that, even if the study shows what the authors claim it shows, so what? They haven’t shown evidence of long-lasting neurological impact, and they certainly haven’t shown any evidence that the hepatitis B vaccine causes autism, even though you know that’s the subtext of what they are arguing. Moreover, the numbers are really small. Another reason is that no statistical justification is given for pooling the no vaccine group with the saline placebo group. Whenever I see this, I become very suspicious of a post hoc combining of data. A good rule of thumb is that it’s usually a bit questionable to combine groups like this. Presumably the reason why two control groups were done was to determine if simply the pain of giving an injection may have had any affects on these neurodevelopmental parameters, a scientifically legitimate reason. But, again, it really makes me wonder that the investigators pooled the data, apparently post hoc. The only reason to do that is to convert three groups to two groups and to add statistical power to the control group. You can be quite sure that, had there been a statistically significant difference between the “vaccinated” group and the saline placebo group and between the “vaccinated” group and the uninjected group, no pooling of data would have been done. Count on it. The results would have been reported un–shall we say?–massaged.

It’s also rather instructive to look at the original IMFAR abstract, which reported:

Kaplan-Meier survival analyses revealed significant differences between exposed and unexposed animals, with delayed acquisition of root, suck, clasp hand, and clasp foot reflexes. Interaction models examined possible relationships between time-to-acquisition of reflexes, exposure, [3C]DPN binding, and volume. Statistically significant interactions between exposure and time-to-acquisition of reflex on overall levels of binding at T1 and T2 were observed for all 18 reflexes. For all but one (snout), this involved a mean increase in time-to-acquisition of the reflex for exposed animals.

It’s interesting to note that the looked at 18 reflexes then but only reported 13 now. Why did they drop five? In this one, there were more “significant” differences in the old group, and only two of the reflexes appeared to be consistent between the two studies. Again, is the experiment reported in this paper a true repeat of the studies in the IMFAR abstracts, or is it simply an “extended” version of the prior study?

I think you know which one I suspect. I also think that this study was a horrible waste of primates, and I can’t believe the University of Pittsburgh’s IACUC was thinking when it approved this study.

On a different note, I wonder about the ostensible justification for this study:

Since Th[imerosal]-containing vaccines, including the neonatal HB vaccine, continue to be used routinely in developing countries [3], continued safety testing is important, particularly for premature and low-birth-weight neonates.

If the authors are so concerned with developing countries, why on earth did they emulate the U.S. vaccination schedule? Why did they use monovalent hepatitis B vaccine, when few countries other than the U.S. do? Most developing countries use a tetravalent, pentavalent, or hexavalent vaccine containing multiple other antigens, such as diphtheria, tetanus, pertussis, IPV, Hib, and HepB antigens. The hepatitis B vaccine, if given at all, usually isn’t given until at least six weeks of age as part of existing vaccine programs. So, when you come right down to it, this study isn’t even looking at what it claims to be looking at or following the rationale that its authors claim. If it were, it would not be following the U.S. vaccination schedule. In reality, it looks very much as though this study is custom-designed to sow doubt and fear about the birth dose of hepatitis B vaccine in the United States. That, and it’s almost certainly going to be used as ammunition for legal action and lawsuits. Just wait.

It also may be another objective here. I note that anti-vaccine groups like TACA funded this study, which certainly cost at least $100,000 to do, probably considerably more. They would not have invested the money if they didn’t expect a payoff. Here’s what I think may be going on. Like all good denialists, they want material to sow doubt about the science they deny. Small preliminary studies in general have a fairly high likelihood of producing false positive results; so funding such studies is likely to produce at least some apparent “hits,” such as this one. Because such studies are small and preliminary, they can’t really settle anything, and the anti-vaccine movement knows it. So anti-vaccine groups like TACA and Generation Rescue will use the results of such small studies as justification for claiming that vaccines are not safe and that money is needed to do bigger studies. They’ll then get such larger studies funded through the NIH. In the meantime, they’ll point to such NIH-funded studies as “evidence” that there is a scientific controversy and milk them for all they’re worth until the larger studies come back negative, as they almost always do.

It’s a very hard strategy to counter, and, unfortunately, it just might work.

REFERENCE:

Hewitson, L., Houser, L., Stott, C., Sackett, G., Tomko, J., Atwood, D., Blue, L., White, E., & Wakefield, A. (2009). Delayed Acquisition of Neonatal Reflexes in newborn Primates receiving A Thimerosal-containing HepatitiS B Vaccine: influence of gestational age and Birth weight NeuroToxicology DOI: 10.1016/j.neuro.2009.09.008