I do not think that study shows what you think it shows

Of all the cranks, quacks, antivaccinationists, and pseudoscientists that I’ve encountered (and applied a bit of not-so-Respectful Insolence to) over the years, there are a few who belong in the top tier—or, if you prefer, the bottom tier. They stick out in my memory for a variety of reasons, either through their sheer crankitude on a variety of subjects (such as Mike Adams), sheer persistence on one subject (such as Jake Crosby or any of the denizens of the antivaccine crank blogs Age of Autism or The Thinking Moms’ Revolution), or fame for promoting quackery (Joe Mercola). One of these elite few, however, does something that’s very useful to quacks everywhere and results in tons of Facebook and Twitter memes that seem convincing to people who don’t know that they distort scientific studies to make it seem as though they support some quackery or other. In fact, with his his partner in woo, he forms a not-so-dynamic duo who run an entire website whose sole purpose is to list studies that can be spun to support pseudoscience and quackery and then, of course, to spin some of them to support pseudoscience and quackery.

I’m referring, of course, to Sayer Ji, and his website GreenMedInfo, a website that subverts legitimate research and misrepresents science in order to support “natural medicine.” Examples of the sort of “intellectual firepower” Ji brings to the issue of “natural medicine” include a post in which he described biotech and vaccines as “cannibalism” and another post in which he tries to represent evidence-based medicine as being no more reliable than a coin flip. The list goes on. I guess that the only consolation is that Sayer Ji (and, apparently Gary Null) really, really like me.

Recently, I became aware of a post that appeared on GreenMedInfo that attempted to use a recent study to claim something that the study doesn’t show. It’s a nice little case study of how science is routinely twisted by cranks to serve their own purposes entitled HPV Vaccine Maker’s Study Proves Natural HPV Infection Beneficial, Not Deadly. It’s written by Sayer Ji and, because seemingly Ji can’t drop such a turd on the world without help, Dr. Kelly Brogan. I haven’t actually dealt with Brogan before, as far as I can remember, but if she works with Sayer Ji, that tells me pretty much everything I need to know about her—well, that, and the panoply of misinformation she lays down in her articles for GreenMedInfo. There’s also her website in which she says she’s practicing “holistic women’s health psychiatry” (whatever that means). What it appears to mean is that she embraces all manner of non-evidence-based woo like homeopathy, acupuncture, antivaccine views, anti-GMO views, and pretty much every kind of medical nonsense I’ve come across. Indeed, her website is such a—shall we say?—”target-rich” environment that I suspect I will visit it again. In fact, this particular article implying that “natural” HPV infection is better at preventing cervical cancer than the vaccine—I kid you not!—is also posted on her website. Nice, job getting a co-authorship without even trying.

In any case, Brogan’s and Ji’s article is festooned with a picture of an attractive young woman holding her hand up as though she is a cop stopping traffic. Amusingly, the version of the article on Brogan’s website has a different woman holding her hand out in the “stop” gesture. Be that as it may, the article starts out by asserting:

Behind every vaccine is an assumption. That HPV causes cervical cancer, that cervical cancer causes death, and that a vaccine can effectively interfere with this linear relationship is the assumption to be examined in this article. Cervarix is a vaccine recommended to girls beginning as early as 9 years old, intended to protect against HPV strains 16 and 18 upon completion of a 3 dose series. It is an aluminum-containing product, with notable “immunogenicity”.

Actually, it’s not an “assumption” that certain serotypes of HPV can cause cervical cancer. Nor is it an “assumption” that cervical cancer can cause death. Nor is it an “assumption” that a vaccine can prevent HPV infection. It’s all science. It’s all been shown. I did notice, though, how she managed to slip in an antivaccine trope about “aluminum,” complete with a link to a bunch of scientific articles purporting to demonstrate the dangers of aluminum, in the classic manner of GreenMedInfo: Without context. Similarly, Dr. Brogan cites a paper that seems to suggest that the HPV vaccine isn’t that good at protecting against high-grade squamous intraepithelial lesions, dysplasias. Of course, it’s just one article, a retrospective study, and the authors themselves write:

…our findings are still consistent with the results of prelicensure and follow-up RCTs (reviewed in Lu et al5). Although the QHPV vaccine was shown to be effective (> 90%) against HPV 16/18-associated dysplasia, VE in practice is likely much lower because these types are only responsible for approximately half of HSILs [high-grade squamous intraepithelial lesions] and a quarter of LSILs [low-grade squamous intraepithelial lesions]. Moreover, these high VE estimates were only observed in per-protocol analyses that were typically limited to HPV-naive women who received all three doses.


However, the allocation of vaccination was not random in this study, so confounding and bias because of differential uptake of screening (detection bias) cannot be ruled out. In our population, vaccinated women were more likely to undergo Pap screening after enrollment. The resulting detection bias tends to mask VE [vaccine effectiveness] because transient cervical abnormalities are more likely to be detected among the more screened vaccinated women.

So in other words, the design of the study tended to mask vaccine effectiveness, and a recent systematic review and meta-analysis showed the vaccine to be very effective in preventing infection with the HPV subtypes protected against by Gardasil (the quadrivalent HPV vaccine). In other words, Dr. Brogan cherry picked her data.

Now let’s see her analysis of the current study, a report on the PATRICIA study (Papilloma Trial against Cancer in Young Adults), entitled Risk of newly detected infections and cervical abnormalities in women seropositive for naturally-acquired HPV-16/18 antibodies: analysis of the control arm of PATRICIA. I’ll briefly boil down what the study showed. Basically, Castellsagué et al examined the risk of newly acquired HPV infection and cervical abnormalities and its correlation to antibody levels for HPV-16 anad HPV-18 antibodies in the control arm of the PATRICIA study.

Basically, the investigators took advantage of the intensive followup of the control arm and the size of the trial to ask a question: Does the antibody response that occurs during natural infection result in any protection against newly acquired HPV infections and the development of cervical abnormalities. The study subjects included women aged 15-25 years old with no more than six lifetime sexual partners, who were randomized to receive either Cervarix or the control hepatitis A vaccine. They underwent frequent Pap smears, with a highly defined method of being examined for cervical changes. Cervical samples and biopsy materials were also tested for HPV DNA and serum samples collected were tested for antibodies to HPV-16 and HPV-18. Endpoints included incident infection (which could include newly acquired infections or recurrent infections present below detection levels at baseline), 6- and 12-month persistent infection, atypical squamous cell of undetermined significance or greater (ASC-US+), cervical intraepithelial neoplasia (CIN) grade 1 or greater (CIN1+), and CIN grade 2 or greater (CIN2+) associated with HPV-16 or HPV-18. (The various CINs are abnormalities of increasing severity between normal cells in the cervix and fully invasive cancer. The higher the grade, the worse the changes.) The analysis included only women who had potentially been exposed to HPV infection via sexual intercourse and those without current HPV infections at baseline. This was done to make sure that only newly detected infections after vaccination were analyzed. A total of 8,193 (1,246 HPV-16-seropositive [15.2%] and 6,947 HPV-16-seronegative [84.8%]) and 8,463 (916 HPV-18-seropositive [10.8%] and 7,547 HPV-18-seronegative [89.2%]) women were included in the analysis.

The authors found:

High titers of naturally-acquired HPV-16 antibodies and/or linear trend for increasing antibody levels were significantly associated with lower risk of incident and persistent infection, atypical squamous cells of undetermined significance or greater (ASC-US+), and cervical intraepithelial neoplasia grade 1, 2 or greater (CIN1+, CIN2+). For HPV-18, while seropositivity was associated with lower risk of ASC-US+ and CIN1+, no association between naturally-acquired antibodies and infection was demonstrated. Naturally-acquired HPV-16 antibody levels of 371 (95% confidence interval: 42–794), 204 (129–480) and 480 (250–5756) EU/mL were associated with 90% reduction of incident infection, 6-month persistent infection, and ASC-US+, respectively.

What does this mean? Basically, it means that the women in the study developed an antibody response to naturally acquired HPV-16 and/or HPV-18 infection, and in some of these women the antibody response appeared to protect against the precancerous cervical lesions caused by HPV infection. Specifically, the study results indicated that antibodies from natural infection against HPV-16 lowered the risk for a newly detected infection (i.e., re-infection with HPV of the matching subtype), atypical squamous cells of undetermined significance or greater (ASC-US+), and cervical intraepithelial neoplasia grade 1, 2 or greater (CIN1+, CIN2+). Moreover, this protection exhibited a dose-response effect. In other words, the higher the natural antibody level, the lower the risk of reinfection. The effect for HPV-18 was much less.

Hilariously, Dr. Brogan objected to the use of the hepatitis A vaccine as a control, intoning that because “there are no vaccine studies in existence using a true non-vaccinated control group, the natural incidence of a disease, as well as the true risks of a vaccine cannot be effectively assessed.” I just mention that because it’s yet more evidence of her and Ji’s antivaccine views. If both groups are vaccinated, even if the vaccines are entirely different, then the study is crap to them. More tellingly, Ji and Brogan (or is it just Dr. Brogan?) attacks a massive straw man:

Given that the entire justification for vaccination is based on the observation that surviving natural exposures to infectious challenges results in lasting immunity, this finding is not that surprising. HPV is no doubt one of countless infectious challenges the body’s elaborate and highly effective adaptive immune system countermands, often subclincally/asymptomatically[1] . Even the authors of the study acknowledged: “Naturally acquired antibodies can remain detectable for at least 4 to 5 years, albeit at much lower levels than those induced by vaccination.” Given the outcomes demonstrating protective efficacy of these lower levels of antibodies, questions may be raised about other elements of the immune system at play in successful pathogenic defense.[2]

The stupid, it burns. Notice the verbal prestidigitation. Brogan and Ji admit that “natural” infection results in antibody production at a much lower level than vaccination does. Yet they still try to imply that this observation, that these antibodies when produced, can be productive. Think of it this way. Only 15% of women continued to have a detectable antibody response to HPV-16 and only 11% to HPV-18. In those women, if you look at the graphs and tables in the paper, you’ll see that the women who were seropositive for HPV-16 had a relative risk of a new HPV-16 of only 0.67 (a decrease by about a third). In other words, only 15% of these women had an antibody response, and those who did only had their risk of reinfection decreased by one third. That’s the best result. For HPV-18, the protective effect of “natural” antibodies against reinfection barely registered.

Once again, Brogan and Ji are attacking the proverbial straw man. The justification for vaccination might be that surviving naturally acquired infectious disease can lead to long-lasting immunity, but the mechanisms by which that happens are different for different diseases, and antibody levels don’t always correlate closely with immunity. This is because other mechanisms might be at work, such as cell-mediated immunity, which doesn’t rely on antibodies the way that antibody-mediated immunity does. None of this stops Dr. Brogan from trying to twist the results of this study into some sort of repudiation of our understanding of immunology and implication that “natural” HPV infection is better than the vaccine, even though the vaccine produces antibody titers that are many-fold higher than even the highest levels of “naturally” acquired antibodies.

None of this stops the obfuscation:

The authors reference the acquisition of identifiable antibodies in only 50-70% of women infected with HPV-16 or 18. What is happening in the rest of the cases? The authors suggest that those with previous infection but without antibodies may have mounted a cell-mediated response that also conferred protection not assessed in this study (or acknowledge to be relevant by vaccine manufacturers). This is tacit acknowledgement of the biochemical individuality that underlies immune response, rendering a reductionist one-size-fits-all model inappropriate for preventive medicine.

OF course, these antibody responses tend to be transient. Moreover, this study shows that it’s the women in which these antibodies persist who are at lower risk for subsequent reinfection and cervical changes associated with the development of cervical cancer. The women who were not seropositive in this study were at the highest risk for reinfection by HPV-16 and HPV-18. Contrary to Brogan’s and Ji’s distortions, this study is actually pretty good evidence that antibody titers, at least for HPV infection, are a pretty good surrogate for immunity; yet Brogan and Ji appear to be arguing exactly the opposite. True, no correlation is perfect, but a dose-response curve in which the risk of HPV reinfection with the relevant subtype as well as the known complications of HPV infection falls as the antibody titer increases is pretty good evidence that antibody levels are important, at least for HPV. Also, it’s not uncommon that natural infection can result in immunity; that doesn’t mean that getting the disease “naturally” is better than preventing it with a vaccine. It is, however, a common antivaccine argument.

There is a grain of a point in Brogan and Ji’s paper, one that’s discussed in the paper they try to misuse. Specifically, it’s the assertion that natural infection can result in immune responses against multiple other types of HPV that can cause cervical cancer. Remember, HPV-16 and -18 only cause about half of the cases of cervical cancer. Still, this is a distinction without a difference. Vaccine manufacturers are already hard at work trying to include more HPV types in new vaccines. Moreover, contrary to the implication made by Brogan and Ji, if there’s anything this study shows it’s that natural immunity for HPV is not something to be relied on to prevent reinfection. Indeed, Brogan and Ji even admit this, but then they go on to write something this dumb:

Given the possibility that HPV infection itself confers protection against HPV reinfection and its more serious health implications (e.g. CIN1+), and that the HPV vaccine may not offer the same level of protection that our inborn adaptive immune responses to natural HPV infection does, we should pause to question the dominant HPV memes circulating out there – via the CDC, the pediatricians, the mainstream media – that proffer HPV is a ‘vaccine-preventable’ disease that we are morally and socially obligated to engage ‘preventively’ through vaccinating every susceptible adolescent on the planet. This fear-borne meme driving the herd to vaccinate against HPV—and everything else —is at least as infectious as the very virus they falsely promise their bivalent and quadrivalent vaccines will defeat.

This is the usual antivaccine piffle. HPV is a vaccine-preventable disease. It makes sense in many populations to vaccinate against it. HPV vaccination itself is pretty crappy at preventing reinfection—according to the results of the very study that Ji and Brogan themselves try to use to claim that we shouldn’t vaccinated against HPV. Truly, in the antivaccine world of Brogan and Ji, up is down, left is right, right is wrong, and we have always been at war with Eastasia—and apparently a study looking at the correlation between antibody levels and reinfection tells us that vaccinating against a first infection is inferior to “natural infection.” Of course, if our immunity to a primary infection were so great, we wouldn’t need a vaccine in the first place. Either Ji and Brogan are incredibly ignorant, or they’re intentionally obfuscating. Take your pick. Neither possibility reflects well on them.

Sadly, Sayer Ji is apparently none too good at handling criticism. There was a meme designed to respond to Ji and Brogan’s torturing of this particular study into submission to the anti-HPV vaccine cause that pointed out the main point that I discussed here (that the study doesn’t show what Ji and Brogan claim that it does) and pointing out how deceptive Ji and Brogan’s article is. I’ve learned from the creators of that meme that Ji has been threatening legal action for defamation, leading them to take the meme down rather than put up with the irritation.

Bad analysis, bad science, and legal bullying from supporters of “natural” healing? Same as it ever was.