The HPV Vaccine (Gardasil) Safety Revisited

Gardasil (qHPV) was licensed in 2006 as a vaccine against four types of Human Papillomavirus (HPV) and marketed as, “The first vaccine targeted to prevent cancer.”  From its inception it has been one of the more controversial vaccines.  Some religious groups feared that the reduced threat of a sexually transmitted disease would lead to increased sexual promiscuity.  Other groups were concerned about its safety.  Some have questioned whether its high financial cost would make it a cost-effective intervention, while others have questioned the marketing tactics of its manufacturer Merck.

Most of these concerns bear consideration (though I have no time for those who advocate using the threat of disease and death to force conformation to their religious beliefs), and were in large part addressed by David Gorski in a SBM article last year.  If you’ve not read his post, I strongly suggest you do so.  Now that a large post-licensure study on qHPV has been published, it seems a good time to revisit the issue of greatest concern to me as a pediatrician and to most parents, namely qHPV’s safety and efficacy profile.


Infection with HPV is a common problem, infecting 79% of the population over their lifetime, and holding the dubious honor of being the most common sexually transmitted disease. In addition to genital warts, HPV infection has been clearly shown to be responsible for a large percentage of cancers in areas that contract HPV, including the vulva, vagina, and penis (40%), anus (90%), and oropharynx (12%), but the most common site of HPV-caused cancer is the cervix.  Cervical cancer is diagnosed in 7.9/100,000 women per year, amounting to approximately 11,000 cases in 2008 and 3900 deaths in the US.  Worldwide, the WHO estimates 500,000 cases of cervical cancer and 260,000 related deaths.

We have largely failed in preventing HPV infection.  Infections are not always obvious and are thus hard for sexual partners to identify and avoid.  Barrier devices like condoms don’t cover all the potential infected areas (see the list above) and are thus helpful but of limited utility. Education and abstinence are clearly inadequate to fully address the problem.

In spite of our limited ability to prevent HPV infection, gynecologists have made significant progress in reducing the number of women who suffer from cervical cancer, in large part through regular screening with Pap-smears.  The ability to identify and treat pre-cancerous lesions on the cervix or to identify cervical cancer early in its course has lead to a marked reduction in deaths from cervical cancer in the US.  It has been so effective that one of the highest risk factors for developing cervical cancer is now a failure to obtain routine Pap-smears.

Yet in spite of the success of Pap-smear screening, we were still faced with 11,000 cases of cervical cancer and 3900 deaths last year.  Through sustained effort and eternal vigilance on the part of patients and physicians we can reduce the impact of HPV induced malignancy, but the population at risk remains unchanged.  And let’s be honest, humans suck at eternal vigilance.  If we could prevent the infection, we could prevent its complications and reduce the impact of the inevitable lapse in our eternal vigilance.  Vaccines allow us to do exactly that.


qHPV is a vaccine against the two most common causes of genital warts (types 6 and 11) and the two leading causes of cervical cancer (types 16 and 18).  Its antigen is a virus-like particle with no genetic material, so it is not a live-virus vaccine and cannot cause infection.  It has an excellent response rate, inducing an adequate antibody response to all four types of HPV in 99.5% of women after the third dose that has been shown to persist for at least five years.

Regarding efficacy, most cervical cancer occurs decades after the infection, so it will be quite some time before we can demonstrate the effect qHPV should have on the incidence of cervical cancer.  However, lesions identified on Pap smear and known to precede cervical cancer occur far earlier, and have served as a surrogate in pre and post-licensure studies.  In trials to date, qHPV is 94-100% effective at preventing pre-cancerous changes from these four HPV types.


qHPV is effective, but what of its safety? When it was submitted for licensure, over 20,000 women between the ages of 16 and 26 (the peak ages during which most women are infected) were enrolled in phase 3 trials.  Potential adverse events were solicited from all participants were nearly identical between the test and control groups.  The only reactions that stood out when compared to the control group were fever (13% vs 11.2%), nausea (6.7% vs 6.5%) and injection site reactions (2.2% vs 0.9%).  No serious reactions were seen.  Based upon this information, qHPV was considered to be safe and approved by the FDA.

Since its release, however, there have been reports of serious adverse reactions following qHPV injection.  Many articles voicing concerns about its safety cite the VAERS database. These reports have generated headlines in the mainstream media, not to mention being perpetual fodder for crank sites like NaturalNews and, and have been a if not the major source of the concern expressed by parents.

As a part of routine post-licensure follow-up and to address these very concerns, “Postlicensure Safety Surveillance for HPV Vaccine” was funded by the FDA and CDC and published last month in JAMA. This article is the largest attempt to evaluate the adverse events seen since qHPV was released.  In it, the authors compared the adverse events reported to VAERS to the expected background rate of reported events in the unvaccinated general population of women aged 9-26, re-evaluating the previously reported adverse events known from licensure as well as the reports of more serious events that have gained public attention.  The results are reassuring.

Over the first 2.5 years since its release, 23 million doses of qHPV were administered, and VAERS received 12,424 reports of adverse events.  Of these, only 772 (6.2%) were classified as serious or life-threatening. These are summarized in the table below, taken from the article:

VAERS reported adverse events for Gardasil

Of all of these events, only two are above what one should expect from their respective baseline rates and not previously identified by the original licensure studies.  The first is Venous Thromboembolic Events (VTEs).  These are blood clots within the blood vessels that can be quite serious; these occurred in 0.2/100,000 doses.  The second was syncope, or loss of consciousness, which occurred in 8.2/100,000 doses.  This does not establish that qHPV causes these events; remember that this study and the VAERS in general are not designed to establish causation.  Nevertheless, further studies are certainly warranted to confirm or refute qHPV’s role in these rare events, and physicians should bear the correlation in mind when considering the administration of qHPV.

Notice, however, that the rates of major events of concern, namely Guillain-Barre syndrome, autoimmune disorders, transverse myelitis, and death, were all exceedingly rare, and not above what one would expect to occur in the normal unvaccinated population.  In spite of the clear limitations inherent in the use of the VAERS database, this study should strongly reinforce the confidence of physicians and parents regarding the safety of HPV vaccination.

Gardasil (qHPV) is a new vaccine.  Recommendations regarding its use are likely to continue to change as we learn to optimize its use.  Just last week, an FDA advisory panel recommended that qHPV be approved for use in males age 9-26, and Merck is likely to seek approval to broaden the approved ages of administration.  Whether these developments should alter the current practice is a point of open debate.  And of course, as with any new medical intervention qHPV warrants close surveillance for any unexpected adverse effects. Though uncertainties remain regarding this vaccine’s use in the sphere of public health, for a parent who desires to reduce the risk of HPV infection and HPV induced malignancies for their child, we can confidently say that qHPV is a safe and effective option.

Posted in: Public Health, Vaccines

Leave a Comment (12) ↓

12 thoughts on “The HPV Vaccine (Gardasil) Safety Revisited

  1. Hi – thanks for the review. I have been following this also, and thinking about it. I believe that a couple realistic diemsions need to be added:

    Evidence sez Gardasil protects against some, but not all, cervical-cancer-causing strains of HPV. Your outcome metrics are not valid. If a 20-year-old female receives gardasil, and receives its protection as you note across years, **but continues to expose herself (no double-entendre intended) to HPV, by repeating the same behaviors, i.e., sexual contact with another person, she will eventually contract a strain of cancer-causing HPV that is not deterred by Gardasil.

    To assess the true likelihood of Gardasil success in preventing cerv CA, you need to divvy up the young female pop into groups based upon epidemiology of HPV and based upon sexual activity from some representative survey. To over-simplify, a young female who has two new sexual partners per year, and if each of those have similarly had about 2 new sexual aprtners per year, such a young adult will eventually contract a HPV strain not deterred by Gardasil. A woman who got Gardasil shot, then has only had sex with her HPV-negative husband, but gets raped by a HPV-pos stranger once, will be at risk for HPV only that one time. That woman is the woman covered by your analysis.

    In my opinion, this is the only honest way to assess benefit. From what I know of sexual expereicen epidemiology, our community will not reap antything close to the the suggested prevention benefit noted by these traditional-but-limited analyses.

  2. nitpicking says:

    MedsVsTherapy, your reasoning is perfect, except that “eventually” can easily be “long after the young woman has died”. People don’t have infinite time to have new sexual partners! If she’s sexually active and non-monogamous for 30 years (to simplify) and has two different partners per year, that’s 60 partners. It’s perfectly plausible that none of these 60 men would be infected by a strain of HPV that’s not prevented by the vaccine. (We also don’t know how effective the vaccine is.)

    Note that most women surveyed seem to have WAY fewer than 60 partners in their entire lives.

  3. qetzal says:

    On top of that, even if the hypothetical women does contract a different strain of HPV, that doesn’t mean she’s now just as likely get cervical cancer. IIRC, different strains have very different tendencies to promote cancer. The vaccine provides immunity to the strains that cause the most cancer. It’s possible that other strains will ‘fill in the gap,’ but I think it’s far from a given (as MedsVsTherapy seems to be arguing).

  4. markentel says:

    I understand that cervical cancer due to HPV is much more common than other cancers (penile, anal) and so is rightly the target of intervention. I am curious as to whether a broader vaccination program including males would provide a beneficial effect?

  5. Harriet Hall says:


    HPV vaccine has already been recommended for males. See

  6. Eric Jackson says:

    Part of the confusion about this seems to be the variety of HPV types available. The CDC quotes the number as more than 130 different types, the National Cancer Institute as more than 100. Some sources up to 200, depending on how finely the types are split up. The CDC considers some 30-40 of these anogenital viruses, or STDs. Of those 40, two(6 and 11) cause the majority(>90%) of visible papillomas. There’s approximately a dozen ‘high risk’ types of the virus which are considered oncogenic, but of these, HPV 16 and 18 are believed to be responsible for >70% of cancer cases, some of the rest by HPV 31, and the other numbers are picking up the scraps left over.

    Further confusing the issue is the prevalence of HPV – complicated by studies that measure only certain types, use different detection methods and count only current confirmed infection, or count prior infection.

    One of the things that doesn’t seem to be mentioned very often is how similar Gardasil is to the HBV vaccine, now almost 30 years old. In both vaccines, the relevant proteins (HBsAg in the case of Engerix/Recombivax, L1 of HPV types 6, 11, 16, 18 in the case of Gardasil) are cloned into S. cerevisiae cells, which are then propagated . The resulting proteins purified to make the vaccine. The meaningful difference between them is only the antigen presented by the vaccine.

    Granted, there’s some slight differences in preparation, adjuvants used. The L1 protein, unlike HBsAg seems capable of self-assembly into a viral capsid. HBsAg isn’t the only component of HBV, and is incapable of doing this. On the grand scale of things though, these two vaccines are almost identical twins.

    The main issue seems to still be the cost. qHPV is still expensive, the CDC selling it discounted at $102 per dose, with three doses required. That’s a hefty price when you’re looking at vaccinating a few tens of millions of people. Hopefully, the price will come down to something more like Engerix – $9 a dose.

    Two things seem to have made this vaccine take up such a big place in the spotlight. One is Merck’s rabidly aggressive advertising campaign. They may end up shooting themselves in the foot with it, but they’re perfectly entitled to do so.

    The other is that it’s against an STD. This has attracted the ire of certain groups active in American politics that take a rather dim view of anything to do with sexual health. The ultimate progenitor of a lot of this controversy seems to be a document put forth by a group called Judicial Watch, which trumpets the use of the FOIA to access VAERS data (which I seem to be able to download here without any sort of legal mumbo jumbo: ).

    The document, located on their rather paranoid Gardasil site here: is pretty biased, inaccurate, and fearmongering. There’s an attempt to spin the acquired data as though the vaccine is causing some massive rash of deaths and horrible consequences, that’s being covered up by the FDA/Government/Merck. The use of the FOIA is prominent and frequent.

    A look at Judicial Watch’s litigation page shows a pretty clear agenda – using lawsuits to harass any politician to the left of Mussolini, prolific litigation on anything to do with illegal immigration, and a surprising number of attacks on anything to do with sexual health. There are lawsuits against Planned Parenthood, against the FDA over the morning after pill, RU-486. There also seems to be a lot of traffic directed their way by religious-right affiliated groups. In short, they’re pretty agenda driven.

  7. brit50 says:

    As the parent of a previously very healthy 20 yr old, who after the second dose of Gardasil in July 2009 developed some very unusual and distressing symptoms: numbness and tingling in the arm where the shot was administered, which moved to both arms, then to both feet and legs. That developed into pain along with numbness and tingling in both legs on exertion.
    We are both puzzled and very concerned to say the least, and worried as to what the future holds. I don’t read those extreme websites mentioned, because I can tell real science from pseudo science. Pouring through the VAERS database reveals many similar reactions, but not the final outcomes unfortunately. A rare condition called Macrophagic Mysofasciitis or MMF has been studied and noted by French researchers, which causes muscle pains on exertion,particularly in the legs, occurring after vaccines containing aluminum salts. This new muscle myopathic syndrome is linked to vaccines containing aluminum, mostly aluminum hydroxide, but also aluminum phosphate which is contained as an adjuvant in Gardasil. From my vantage point, I must join the ranks of people who question the safety of Gardasil. I am wholeheartedly in favor of vaccines, including one against HPV, but this event has give me much reason to pause. The use of adjuvants such as aluminum salts, which have been used since the 1930’s and are considered “relatively safe” is something I would like to see discussed more in the scientific community. There are even more adjuvants used in Europe which have not been approved in the US as the FDA is not convinced of their safety. I am grateful for this caution on the part of the FDA . I believe there is a need for more adjuvants with better safety profiles in order to continue to provide safe vaccines for the public. I would like to see more research in this area. I wouldn’t wish for one single person to go through what we’re going through.

  8. brit50, as sad as your story is, please explain to us what evidence you have that shows there is any link between the vaccination and Gardasil? And what evidence do you have that your daughter’s condition may not have been caused by anything from a psychological reaction to an allergy to one of the components?

    Did you not read the article? Seven hundred serious adverse events in 22 million doses, and some of those adverse events are not statistically different than what is expected in an non-immunized population.

    It’s easy to blame something, but it’s much harder to show a scientific cause and effect. That’s because science has standards way above anecdotal.

  9. Squillo says:

    Thanks for this thorough discussion.

    Stupid layman question:

    Is it possible that the excess episodes of fainting associated with HPV vaccine administration is due to previously undiagnosed vasovagal syncope?

    The age for administration of the vaccine seems to overlap with the average age of onset.

Comments are closed.