The real Sasquatch genome scandal

SmalfutThe claim that a group of researchers, led by Melba Ketchum, had sequenced the Sasquatch genome was one of the wackier science stories of 2013.

Unable to find a home for their findings in any established journals, the authors started their own journal–DeNovo–and published their article on the Sasquatch genome there. (So far, this is the only article that has been published in the journal.) They held a press conference in Dallas on October 1st, and the media had a lot of fun covering the story.

To their credit, Ketchum and co-authors have made their article and the supporting data public. You can check out the article here and the supporting data here. They even made available some peer reviews that had been leaked after they submitted their work to Nature and the Journal of Advanced Zoological Exploration in Zoology (I’m not familiar with the latter journal). They also seem to have been very open in discussing their work with other scientists, which is great.

John Timmer provided a great summary of the considerable problems with the Sasquatch study over at Ars Tehnica. Here’s the short version: the sequence produced did not belong to bigfoot. It was an artifact resulting from sample contamination, degradation, and subpar assembly methods.

So OK, a wacky bit of research comes out, it’s rejected during the peer review process, and the post-publication process seems to have worked well too. The media seems to have accepted that this is not solid science, although it made for some great headlines. The system worked, and no real harm was done. Right?

That’s what I thought initially, but in reading about this story something started nagging me. Ketchum is a forensic scientist. She is director of a company called DNA Diagnostics that appears to specialize in animal forensics. Forensics is all about quality control, correct? If elementary quality control measures were not taken in this bigfoot study (and don’t even seem to have been understood), what does that say about the forensic work of this team–and in particular the work of the lead author?

This study was quickly debunked by scientists because it was published, albeit in a vanity journal, to great media fanfare. But forensic evidence is usually presented in court absent any real peer-review process. As the ongoing Annie Dookhan saga illustrates, when the integrity of forensic evidence is compromised, it can have real consequences. People go to prison for crimes they didn’t commit. Other people don’t go to prison because the evidence that should have put them behind bars is thrown out. Murders go unsolved. This is serious stuff.

According to her CV, Ketchum has presented forensic evidence in criminal cases. In death penalty cases, even. Death penalty cases in Texas, which has executed more prisoners since the death penalty was reinstated in 1976 than any other state. If basic quality control procedures were neglected in the presentation of this criminal evidence, as they were in the bigfoot study, what does it mean for the outcomes of the cases involved? (Incidentally, the problems with DNA Diagnostics–the lab Ketchum runs–seem not to be limited to bigfoot research.) What does this say about the quality of scientific evidence being presented in life-or-death cases in general? This reminds me of the incredible problems with expert scientific testimony revealed for shaken-baby syndrome and arson. I found the whole thing very, very unsettling.

A side effect of the flu vaccine teaches us something new about narcolepsy

images-3You’ve heard of narcolepsy. Probably the most famous symptom is someone falling asleep during the course of normal, daytime activities (like conversations). It’s a neurological disorder in which the brain is unable to properly regulate sleep-wake cycles. Experiencing sudden muscle weakness when you feel strong emotions is another symptom, as is sleep paralysis (when you wake up but cannot move or talk). Reading these symptoms makes me wonder if I have it! But it’s rare–it has been estimated to affect about 2 out of every 10,000 people.

We don’t have a great handle on what causes narcolepsy. Over the past few decades, evidence has started mounting in favor of a new hypothesis: that narcolepsy is an autoimmune disease. It became clear that the disease is  associated with a certain HLA genotype, DQBq*0602 positive (the human leukocyte antigen or HLA genes are involved in immunity). About 30% of the general population carries this genotype, but it’s found in 99% of narcolepsy cases, implicating the immune system. In addition, when researchers injected antibodies from narcoleptic humans into mice, the mice developed narcolepsy-like symptoms.

Now we’ve got a new source of evidence supporting the autoimmune hypothesis: a link between a flu vaccine used in 2009 and a scattering of childhood narcolepsy cases. As you may remember, 2009 was the year of the big H191 flu scare. In Europe, some countries used a vaccine called Pandemrix that was produced by GlaxoSmithKline. Researchers in Scandinavia were the first to discover that getting the flu shot was a risk factor for developing narcolepsy. In western Sweden, the odds of developing narcolepsy were roughly 25x greater for kids that got the Pandemrix shot than those who did not; in Finland, they were 13x greater; and in England, they were about 16x greater. Eventually, similar results were found in children who had received the vaccine in France too. Here, you can see a graph depicting the spike in cases following the 2009 vaccination campaign in Finland. Pretty striking!


Spike in childhood narcolepsy cases in Finland following H1N1 vaccination. From: Partinen M, Saarenpää-Heikkilä O, Ilveskoski I, Hublin C, Linna M, et al. (2012) Increased Incidence and Clinical Picture of Childhood Narcolepsy following the 2009 H1N1 Pandemic Vaccination Campaign in Finland. PLoS ONE 7(3): e33723. doi:10.1371/journal.pone.0033723

Now this is scary stuff, but I’d  like to point out that narcolepsy is actually a rare side effect of the shot. Only 1 of every 16,000 vaccinated children developed narcolepsy. And it’s important to remember the H1N flu itself can be pretty nasty, which is why the vaccine was developed in the first place. In Sweden, for example, even with the vaccination campaign there were 4,753 cases of H1N1 flu in children: 571 needed hospital care, 27 needed intensive care, and 4 died. So death occurs in about 1 of 500 kids who get the flu ,and narcolepsy occurs in about 1 of 16,000 kids who get vaccinated with this particular shot. Balancing the risk associated with the shot and the risk associated with the flu is tough–but for my kids, I would still favor the odds on the shot. Nevertheless, by 2011 the WHO had restricted the use of Pandemrix in children. This makes sense since other types of flu shots — that have not been linked to narcolepsy — are available.

Because narcolepsy seemed to be a side effect primarily of the European vaccine, attention centered on an ingredient found only in Pandemrix. This vaccine contained an adjuvant called ASO3 (an adjuvant is a material that helps create a strong immune response), whereas US versions did not. Thus, scientists hypothesized that it could be the strength of the immune response, which is greater when an adjuvant is involved, that leads to increased narcolepsy risk.

Now here is where things get tricky. It seems pretty clear at this point that in rare cases, the Pandemrix H1N1 vaccine triggers narcolepsy. But it looks as though the flu itself can trigger narcolepsy too. Getting the H1N1 flu, in particular, seems to be an important risk factor. Researchers at the People’s Hospital of Beijing University compared the number of new narcolepsy cases diagnosed every year from 1998–2012. They found a huge spike in new cases in 2010, following the H1N1 winter flu epidemic there. In 2010, 201 new cases were identified–more than the 187 new cases identified in all of 1998–2009 combined! By 2011 and 2012 the number of cases were back down to normal. So it looks like both the vaccine AND the flu can lead to narcolepsy. Doesn’t this remind you of the post I wrote about the flu vaccine and Guillain-Barré syndrome a few weeks ago?

The kind of risk associated with flu is not entirely clear yet, however, as H1N1 infection doesn’t seem to have contributed to the narcolepsy spike in Finland or in some other countries. Perhaps infection was more intense in China, in terms of both the numbers of infected people and the strength of the immune response. It’s not clear yet, but it will be interesting to see what we learn as time goes by.

So why would the flu vaccine/flu cause narcolepsy? In a recent article in Science Translational Medicine, researchers found that patients with narcolepsy have immune cells primed to attack a hormone called hypocretin. The brains of people with narcolepsy are missing the neurons that produce hypocretin, so this makes sense. Their immune system appears to have killed them off. And here’s where it all comes together: part of the H1N1 hemagglutinin protein closely resembles the parts of hypocretin visible to immune cells. In the laboratory, exposing immune cells from narcolepsy patients to these viral fragments increases the frequency of hypocretin-reactive cells! This is no good for the cells that produce hypocretin and suddenly become attractive targets for the immune system.

Who is most at risk for developing narcolepsy after the shot? All post-vaccine cases identified by three big narcolepsy centers in France, Canada, and the US have had a specific HLA genotype: DQBq*0602 positive. That was the genotype I mentioned before, the one that 99% of narcoleptics have. The post-vaccination cases identified in Sweden also had this genotype. So it seems as though the shot isn’t causing narcolepsy in just anyone. Instead, Pandemrix seems to be a precipitating factor for children who are already susceptible to developing the disease. Therefore, researchers have proposed HLA typing individuals prior to administering flu vaccines like Pandemrix.

Researchers have also called for additional studies examining whether non-adjuvant flu vaccines increase the risk of narcolepsy. The H191 vaccination campaign was unusual in that it involved the mass vaccination of tons of kids in a short period of time, facilitating our ability to recognize side-effects, like narcolepsy, linked with the shot. It’s also notable, but not surprising, that the narcolepsy connection was discovered in Scandinavia, where over 75% of kids were immunized and the health reporting is of excellent quality. It’s still possible that other flu vaccines carry a risk, probably less pronounced, that is hard to detect under normal conditions (like the relative chaos that is the US healthcare system). I think we are going to see some interesting research emerge in this area!