Field of Science

Stimulus funds are promoting pseudoscience at Harvard Medical School

Another in an ongoing series of blog posts about wasteful spending by NCCAM, the National Center for Complementary and Alternative Medicine

In today’s competitive research environment, if the government offers money for something, then someone will do it – even if the research is worthless. This even happens at Harvard (where, it so happens, I got my Ph.D.).

Let's look at how, just recently, NIH funnelled more money to acupuncture. Bruce Rosen of the Massachusetts General Hospital and Harvard Medical School has secured two stimulus grants totalling $1.8 million to study acupuncture’s effects on brain activity. This was clever in two ways: first, Rosen took advantage of the fact that the National Center for Complementary and Alternative Medicine (NCCAM), the home of pseudoscience within NIH, had some stimulus funds that it had to spend quickly. Second, his studies simply use imaging techniques (fMRI) to look at patients’ brains during acupuncture. This virtually guarantees that he’ll find something: if you stick a needle into a person’s body, then of course the brain reacts! It hurts, doggone it! And if you compare this to sham acupuncture, where the pain is a bit different, then voila! you find that the fMRI looks different! Ergo, acupuncture must be activating some special brain functions. Clever!

Specifically, here are the stimulus funds that Rosen just landed:
2P01-AT002048-06 “Neuroimaging Acupuncture Affects on Human Brain Activity, $1,200,061
1P30-AT005895-01 , “Core Center for Multimodal Evaluation of Acupuncture Mechanisms”, $593,196

Here’s how Rosen introduces one of these projects:

“Acupuncture, a component of traditional Chinese medicine, has been used for thousands of years to treat a multitude of ailments. Recent scientific evaluation has suggested that this therapy may demonstrate clinical benefit for a number of conditions including chronic pain, though the mechanisms of action have not been well understood.”
Here we have, in his very first sentence: a classic logical fallacy (the “argument from authority”), right at the beginning of the proposal. Even if a treatment has been used for thousands of years (and let me note here that this is not true of acupuncture, despite numerous claims to the contrary), that doesn’t mean that it works. And of course, in ancient times people had a life expectancy of about 30 years, so we don't exactly want to go back to those days. Rosen follows this fallacy with the “suggestion” that acupuncture might work – and note how vague his claim is, using the phrase "may demonstrate" – despite the fact that all the properly-controlled studies show the opposite. In fact, recent studies show quite convincingly that “sham” acupuncture works just as well, even if the needles don’t penetrate the skin! And with placebo treatments, you get the benefit (such as it is) without the risk of infection from needle sticks.

Rosen has impressive credentials, with a Ph.D. in physics from MIT and an M.D. from Hahnemann Medical College. And he’s probably a really smart guy - who got taken in by pseudoscience somewhere along the way. A look at the publications on his acupuncture center’s website reveals almost no publications in the peer-reviewed literature since 2005. OK, maybe his web page is out of date, so I looked on PubMed and found two – only two – publications by Rosen on acupuncture and MRI in the past two years. Both were small, poorly controlled studies, and one appeared in BMC Complementary and Alternative Medicine, a rather pathetic excuse for a journal. With this kind of productivity, no wonder NCCAM gave him another $1.8 million!

But NCCAM loves this stuff. They even feature a picture from one of Rosen’s (two) studies on their website in their latest newsletter. It doesn’t matter if acupuncture works, as long as you can generate cool MRI images of people’s brains. Hey look, it’s an MRI – it must be science!

So Bruce Rosen is getting $1.8 million in stimulus funds to take worthless MRI images of patients’ brains while someone sticks them with needles. And this $1.8 million is merely a supplement to his ongoing NCCAM work – he’s been funded by NCCAM since at least 2003, (see the “06” at the end of that first grant number above? That means it’s in its 6th year of funding) and he proudly states that “our well-established group has been a leading force in acupuncture mechanism research for over a decade.” Great.

Finally, if Bruce Rosen reads this and feels I’m unfairly attacking him, well, he can reply on this blog (comments welcome!), or he can talk to me in person in February, when I’ll be giving a talk at Harvard Medical School, his home turf.

Merck hiding negative results about drugs, again

You’d think that after the Vioxx fiasco, Merck might be a little more careful about pushing drugs that don’t offer much benefit. Well, if you did think that, you’d be wrong. A new study in the New England Journal of Medicine this week shows that two very, very profitable cholesterol drugs are pretty much useless. Merck has been at least partly aware of this for 3 years, but they still don't want to admit it – not surprising given that sales of these drugs by Merck last year totalled over $4.5 billion. That’s “billion” with a "b." Merck has also been advertising these drugs heavily since they were approved by the FDA in 2002, and it has paid off.

So what are these drugs? They’re called Vytorin and Zetia, and first of all, they’re not statins (which many of you may have heard of), which are indeed useful at lowering cholesterol. The newer drug is called ezetimibe; Zetia is the brand name, while Vytorin is a combination of ezetimibe and a statin drug. Merck makes statins too, and they have been enormously profitable, but with statins I’d say the drug makers have earned their profits (at least some of them).

What did the new study find? Well, to quote the article itself, “niacin is superior to ezetimibe”, and the NEJM editors wrote that, “the findings, obtained in a modest sample of 208 patients, followed for only 14 months, show a clear superiority of niacin over ezetimibe.” That’s right, the use of inexpensive, non-prescription niacin (vitamin B) was clearly better that Zetia. It’s worth noting that the study wasn’t measuring the rates of heart attack; instead, it measured the thickness of the common carotid artery as a surrogate, on the assumption that thickening of the arteries leads to heart attacks. But there were also more heart attacks, and more deaths, in the patients on Zetia – not enough to reach statistical significance, but perhaps that was only because of the small number of subjects. The trial was halted early when it became clear that niacin worked better, and it might be harmful to the subjects to continue giving them Zetia (ezetimibe) instead.

I found another NEJM article from April 2008 that showed that combining Zetia with statins was no better than statins alone.

So what did Merck have to say? Astonishingly (or not?), they are continuing to promote Zetia aggressively, and they issued this statement a few days ago: “"The results of the small ARBITER 6 study do not, in any way, change our view of Zetia and Vytorin as effective medicines for fighting high LDL cholesterol," said Peter S. Kim, Ph.D., president, Merck Research Laboratories. “We encourage patients to continue taking their medication as prescribed by their physicians.” Really? The results don't change his view in any way?

This latest news comes just 3 months after Merck and Schering-Plough agreed to pay $41.5 million to settle a lawsuit that accused them of sitting on unfavorable results about Zetia and Vytorin for two years. That lawsuit was based on results released in January 2008, describing a study that ended two years earlier.

This is the kind of greedy, evidence-be-damned behavior that encourages conspiracy theorists to make wild accusations about drug companies. The problem is, sometimes their accusations are true, even if the drug companies do produce many beneficial products. One thing the U.S. should do is put a stop to direct-to-consumer marketing of drugs, ads like the one below, which claims that Vytorin is better than statins alone. Note how it doesn’t directly claim that Vytorin will reduce your risk of heart attacks, but it cleverly implies that it will:

Now that is insidious. We should never have allowed direct marketing of drugs, and we can still put a stop to it.

10,000 genomes – why?

In the genomics world that I inhabit, a consortium has just published an intriguing proposal to sequence the genomes of 10,000 vertebrate species. It’s described an article in the current Journal of Heredity – unusual in that this is not a research article, but a proposal. Nonetheless, the article is full of interesting facts about what we know (and don’t know) about vertebrate species and how they’re all related. It makes a good read for anyone interested in evolution; for example, how many people know that all vertebrates have a common ancestor who lived about 500-600 million years ago? Perhaps more interesting, evidence is emerging that we all share about 10,000 genes – which means that these 10,000 genes are so useful that their functions have been preserved for 500 million years.

The consortium is led by a number of outstanding scientists, including UCSC’s David Haussler and NIH’s Stephen O’Brien, both of whose work I like and have followed for years. And some aspects of this proposal are terrific: for example, they want to start collecting DNA now from 16,203 vertebrate species. This will make a great specimen collection for future work. (Heck, maybe I’ll even sign on to the project myself.)

But this proposal is more than that: it is also the opening salvo in an effort to raise $50-100 million for the sequencing of these species. The paper was announced with press releases and news articles in both Science and Nature, demonstrating that it is clearly a lobbying effort for new funding. Fair enough – science takes funding, and sometimes you have to build support for new ideas. However, given that two of the three leaders of the consortium are primarily funded by NIH, I can only guess at who they’re expecting to cough up the money. The NIH's human genome funding has been led by NHGRI, which continues to look for new ways to justify maintaining the size of its three enormous sequencing centers (the Broad Institute, Washington University in St. Louis, and Baylor College of Medicine). Now, let me say here that genomes are the bread and butter of my own work, and these centers have done terrific work for the past 10-15 years - and I've often collaborated with them. And I hope they will continue. But it hasn’t escaped my attention that the new NIH Director, Francis Collins, was a key force in building up those centers while he was Director of NHGRI, and we all know that the centers are near and dear to him.

But wait a minute: 10,000 genomes, none of which are human? Exciting idea, sure, but not for NIH. The NIH-led centers are already participating in the 1000 Genomes project, which is attempting to sequence 1000 individuals (at a low-level “draft” quality). If NIH wants to scale up, there are 6 billion more of us available. Admittedly, not everyone wants to have his/her genome sequenced, but plenty of us would be happy to volunteer. Take a look at the Personal Genome Project, which was started by George Church at Harvard, and is trying to sign up 100,000 humans to have their genomes sequenced.

So I’m going to be the skeptic here: if NIH is thinking of throwing its sequencing dollars at 10,000 more genomes, I suggest that it focus on humans rather than a broad collection of other vertebrates. (I know, this should be obvious, right? But sometimes NIH gets distracted.) We still have only scratched the surface of what there is to know about the human genome, and there’s plenty of DNA sequencing to do in the human population. The 10,000 genomes project sounds like a great mission for the National Science Foundation, which has ceded much of large-scale sequencing work to the NIH in the past 10 years, except for plant genomes and some bacteria. In fact, maybe this is a chance for NSF to have its own large-scale sequencing center – I’d be all for that. But I’m not at all convinced that NIH should spend its biomedical research dollars on 10,000 vertebrates. Let’s see how this plays out.

Flu vaccine shortage

This is just a brief post to point out a news article on the wires today (from Agence France-Presse) quoting me at some length about the shortage of H1N1 (swine) flu vaccine in the U.S. The article discusses why the egg-based vaccine production method, which is woefully obsolete, needs to be replaced if we're to avoid such shortages. I also made a point about how risky it is for us (as a society) to be relying on chicken eggs for a virus that could wipe out chicken flocks. See the article at Yahoo news, here, or at the Daily Telegraph in the UK, here.