Oh, Discover. You’re such a tease. You have Ed and Carl and Razib and Phil and Sean, an (all-male, ahem) cluster of science bloggy goodness. But then you also fawn over HIV deniers Lynn Margulis and Peter Duesberg. Why can’t you just stick with the science and keep the denial out?*
But no, now they’ve let it spill into their esteemed blogs. I was interested to see a new blog pop up there, The Crux, a group blog “on big ideas in science and how these ideas are playing out in the world. The blog is written by an outstanding group of writer/bloggers and scientist/writers who will bring you the most compelling thoughts throughout the world of science, the stuff most worth knowing.” Sounds ok, let’s see what stories are up…oh, one on HPV! Right up my alley. And hey, a woman! Bonus.
Ohhhhh, it’s actually one on HPV vaccine misinformation, written by the author of the fawning Duesberg article referenced above. Faaantastic.
The author, Jeanne Lenzer, poses the question, Should boys be given the HPV vaccine? and answers herself that “the science is weaker than the marketing.”
What does she base this conclusion on? A recent paper examining HPV vaccine in boys, maybe? No, of course not. She uses the 4-year-old NEJM study demonstrating the efficacy of the then-new vaccine as the main basis of her claim, and it’s a house of cards from there. Let’s look at what she says, shall we?
First, let’s start with the 2007 NEJM paper. This paper tested the multivalent HPV vaccine, which contains HPV types 6 and 11 (these cause warts and aren’t further examined in this particular paper) and 16 and 18 (cause cancer, focus of this research). The goal was to “assess the prevention of cervical intraepithelial neoplasia grade 2 or 3, adenocarcinoma in situ, and cervical cancer caused by HPV-16 or HPV-18.” This makes sense that they’re looking at cancer outcomes only due to types 16 and 18–after all, while other HPV strains can cause cancer, HPV 16 and 18 are the only cancer-causing types included in the vaccine. They even note in the introduction that HPV 16 and 18 cause about 70% of all HPV-related cervical cancers, but that’s still not 100%. I mean, they state this right in the paper: “The primary hypothesis stated that, as compared with placebo, the vaccine would reduce the incidence of high-grade cervical intraepithelial neoplasia related to HPV-16 or HPV-18 in the per-protocol susceptible population.” I spell this out because this is one of Lenzer’s criticisms, which I’ll come to in a bit.
So they have this vaccine. Who did they test it on? In this trial, it was women ages 15-26, and the mean age of this group was approximately 20 years old. Age at sexual debut was 16–so on average, women in this trial had been having sex for about 4 years with 2 sexual partners by the time they were vaccinated. Now, for those of you familiar with the vaccine, that age range may seem a bit odd–because the vaccine is currently recommended to be given to girls starting at around age 11. But you see, this is one way they ended up at that recommendation–because of SCIENCE! (I’ll also come to that in a minute).
All right, they have young women, initially just a bit over 12,000 of them in 13 countries. Let’s look at their eligibility criteria (who could actually participate):
Women were eligible to participate in the study if they were not pregnant, did not report abnormal results on a Papanicolaou smear, and had had a lifetime number of no more than four sex partners.
That’s it, and it’s an important thing to note. For initial enrollment, they didn’t have to be HPV negative. This becomes important later.
The eligible women were randomized into the treatment or placebo groups, and then received 3 doses of the vaccine (or placebo), and were asked to report any serious adverse events for up to 15 days post-vaccination. The women also underwent pap smears and HPV testing at this point. They also had a number of follow-up visits to be re-tested, up to 4 years post-vaccination.
About those baseline tests. Even though they enrolled a bit over 12K women, they found that at baseline, pap smears were abnormal for 11.8% of subjects in the vaccine group and 11.1% in the placebo group. In other words, they were already infected with a strain of HPV at the beginning of the study. They further examined everyone using both DNA and serological testing for HPV 16 and 18, and found that about 10% of their cohort was already positive for HPV 16, and about 3-4% for HPV 18. Thus, they analyzed their data in several different ways, including doing an analysis including the pre-existing HPV 16 & 18 positive participants, and another one without them.
They also analyzed by eliminating participants who hadn’t fully followed or completed the protocol, including getting only one or two doses instead of the targeted three; having missing samples; or “general protocol violations,” which apparently consisted largely of not getting the month 7 swab at the right time, or administration of IgG, immunosuppressive drugs, or blood products–all things that could affect study outcomes.
Now the numbers. When they looked at women who’d followed protocol *and* were susceptible (ie not already infected with HPV 16 or 18 upon entry into the study), efficacy was 98% at preventing HPV-16/18-related high-grade cervical lesions. And what are “high-grade cervical lesions,” you might ask? According to the NCI:
High-grade means that there are more evident changes in the size and shape of the abnormal (precancerous) cells and that the cells look very different from normal cells. HSILs are more severe abnormalities that have a higher likelihood of progressing to cancer. HSILs include lesions with moderate or severe dysplasia or carcinoma in situ.
This is basically the step before cancer, and as noted above, one of the endpoints of the study, yet Lanzer calls them “innocent cellular abnormalities” and cites this paper, which doesn’t match her claim. If she’s referring to the line that “most HPV infections are easily cleared by the immune system,” I hate to break it to her that the ones that are “easily cleared” aren’t usually the ones causing the more serious dysplasia and carcinoma in situ. Rather, those HPV-caused lesions typically are examined via colposcopy and biopsy of the area to check for cancer. (Bonus scavenger hunt for the phrase “coffee ground-like discharge!”). Next, a woman with abnormal cells can expect to undergo a LEEP procedure, where portions of your cervix are removed with a burning electric wire under local anesthetic, and the foul smoking remains of your cells are sucked up into the smoke shark, “a sleep, powerful, smoke-eating machine.” After that one, side effects include infection, hemorrhage and possibly cervical incompetence. These are rare, but if we’re talking vaccine side effects versus possible outcomes from HPV infection, these types of outcomes need to be considered as well–not just death from cervical cancer.
What I’m trying to say here is that Lanzer may be minimizing these “innocent” non-cancer findings just a wee bit.
OK, so 98% efficacy in their target population who followed the study perfectly. What about if they included everyone? For that, they do what’s called an “intent to treat” analysis, basically putting everyone back in the pot. As Lanzer notes, this is when efficacy falls to a reported 44%, but she completely fails to note that the bulk of this was because the women had pre-existing HPV 16/18 infections and/or lesions. Like almost all vaccines, getting it isn’t going to help you if you’re already infected and showing symptoms. As I mentioned before, SCIENCE! This is why it’s so strongly recommended now to get the vaccine prior to the onset of sexual activity–if you’re already HPV-infected, the vaccine will not do you as much good. Hence, the current recommendation to receive the vaccine prior to the onset of sexual activity (ages 11-12ish).
OK, I’ve had about all I can take and this is getting to be a post of Oracian lengths. To jump to the end–why didn’t they just wait 20 years or so to see if there would be more cancers in the placebo versus unvaccinated group? Besides the whole waiting 20+ years thing, maybe because that would be totally unethical? I’ll take this right from the discussion section:
Although prevention of invasive cervical cancer is the main goal of prophylactic HPV vaccination, it is ethically unacceptable to use invasive cancer as the end point in efficacy trials. Cervical intraepithelial neoplasia grade 3 and adenocarcinoma in situ, which the International Federation of Obstetrics and Gynecology classifies as stage 0 noninvasive cervical cancers, are clinically important outcomes because they are likely to persist and may become invasive without treatment.
Thing is, even if they had followed these women for 20 years, doing checkups say every 6 months, they would have to follow standard treatment protocols (like the LEEP nastiness I mentioned above) and as such, would be doing everything they could to prevent cervical cancer in any case. So, these precancer endpoints are more ethical substitutes, and still show the ability of the vaccine to prevent abnormal cell changes that can lead to cancer.
Finally–Lenzer uses her supposed refutation of HPV vaccine efficacy in girls to then argue that we’re looking to give this vaccine to boys unnecessarily. She cites first that there are only 300 anal cancer deaths in the US every year (one of the cancers the HPV vaccine should reduce in men–and women, I might add). Small potatoes–why give the vaccine to boys over 300 deaths, right? But oddly enough (or, not?), she gives the barest of lip service to a group of HPV-associated cancers which affect men more than women, head and neck cancers. Guess what–there’s a lot more of those than anal cancers, and several studies suggest that HPV-associated head and neck cancers are actually increasing in incidence.
Even when she does mention oral cancers, she asks rhetorically, “what about the claim that Gardasil might prevent anal and oral cancers men may get from having sex with other men?” Um….why focus only on men having sex with men? Perhaps it hasn’t occurred to Lenzer that oral and anal cancers in men can also be transmitted to them by female partners, and vice versa–so we’re not just talking about “gay” cancers in men. I wonder if her wording was purposeful, though, as it’s easier for many in society to dismiss concerns about diseases that are transmitted via homosexual sex acts.
To wrap up, some drama. Lenzer frets about the possibility of developing Guillain-BarrÃ© syndrome from the HPV vaccine:
Given this, is it worth the risk of exposing millions of youth to the as yet uncertain harms of the vaccine? The CDC states that in rare instances, some vaccines may trigger the potentially fatal and paralyzing condition Guillain-BarrÃ©, and Nizar Souayah, MD, of the University of Medicine and Dentistry of New Jersey in Newark, says he and his colleagues found “clear evidence from our database of an increased incidence of Guillain-BarrÃ© syndrome in the first six weeks, especially the first two weeks, after [HPV] vaccination.” Guillain-BarrÃ© is very rare, even among people who are HPV vaccinated, but the problem is emblematic of the downsides of subjecting millions of people to any medical treatment.
The link takes you to a WebMD article from the 2009 American Academy of Neurology meeting reporting on a study by Nizar Souayah. The WebMD article states:
Overall, the vaccine does not raise the odds of developing Guillain-Barre syndrome (GBS), a disorder of the peripheral nervous system, says Nizar Souayah, MD, of the University of Medicine and Dentistry of New Jersey in Newark.
“But there is clear evidence from our database [the VAERS database–TS] of an increased incidence of Guillain-Barre syndrome (GBS) in the first six weeks, especially the first two weeks, after vaccination,” he tells WebMD.
Still, the risk is extremely low: 26 in 10 million in the first two weeks and 30 in 10 million in the first six weeks after vaccination. That compares to 5 in 10 million odds in the general population, Souayah says.
Leznar bills herself as an investigative reporter, but didn’t even bother to dig up the published version of this research. The numbers are fairly similar, so even if we accept that your chance of developing GBS from the HPV vaccine are 30 in 10 million (which are a bit sketchy coming from the VAERS database), that would still be only 3-6 times higher than those who didn’t receive the vaccine, with a condition that’s extremely rare to begin with. (To put it in perspective, the likelihood that you’d be struck by lightning in any year are about 1 in 750,000 according to NOAA). You have much better odds of contracting GBS from eating a poorly cooked chicken breast, as Campylobacter infections are a bigger cause of GBS than vaccinations. Even influenza infections cause more GBS than vaccines.
To sum up, in case it’s not painfully obvious by now, Leznar’s article is full of half-truths and omitted facts–so many that it would take me an entire series of articles to go through the rest of her post, and I haven’t even gotten into any post-2007 HPV vaccine research! To top it all off, she doesn’t even try the “fair and balanced” journalist approach, instead quoting only vaccine “skeptics” but not those who are in favor of HPV, and of then ends on the “I’m not anti-vaccine” note–she can’t be, as she thinks smallpox is bad! (Of course you’re not, neither is anti-vaccine advocate Barbara Loe Fisher. Is there a drinking game for that somewhere?)
I’ll end with my own addendum.
My 11-year old daughter is vaccinated for HPV. My 9-year-old son will be vaccinated. I wish they’d had the vaccine when I was younger, so I didn’t have first-hand knowledge of the Smoke Shark, who still haunts my dreams.
*And yes, I know that Discover is owned by the Gucciones, who have ties to another HIV denialist besides Duesberg and Margulis. [UPDATE: Turns out Guccione doesn’t appear to own Discover any longer, and that it was sold last year to Kalmbach publishing. Thanks to Ed & Razib for the correction.]