The reduction of mortality and morbidity attributable to HPV-associated cancers is generally the primary aim of HPV vaccination programs. Many have criticized the wide-scale implementation and recommendations of HPV vaccination programs on the grounds that HPV vaccines have not been proven to prevent any cancers (Abdelmutti and Hoffman-Goetz 2009; Dyer 2015; Lippman et al. 2007; Syrjänen 2010; Tomljenovic and Shaw 2013; Tomljenovic et al. 2013). Critics have argued that the etiological link between HPV infection and cervical cancer is too poorly understood to warrant wide-scale vaccination, given that many women who are infected with HPV never develop cervical cancer, and many precancerous lesions may regress spontaneously (Lippman et al. 2007; Rail and Lippman 2015; Tomljenovic and Shaw 2012b; Tomljenovic and Shaw 2013).

Phase III HPV vaccine clinical trials have shown extremely high prophylactic efficacy against the intermediate endpoints of persistent infections, genital warts, high-grade cervical lesions, high-grade anal lesions, and vulvar/vaginal lesions with vaccine-type HPVs in women and men who were not previously infected with those types (Beachler et al. 2016; Future II Study Group 2007; Garland et al. 2007; Giuliano et al. 2011; Joura et al. 2015; Joura et al. 2007; Paavonen et al. 2009). These trials did not evaluate efficacy against cervical cancer and other cancer outcomes for ethical and practical reasons. Firstly, it would not be ethical to allow precancerous lesions to progress to cancer in participants; lesions are treated upon detection. Secondly, the progression from initial infection to cervical cancer is a decade-long process (Schiffman et al. 2007); it would be unfeasible to prolong trials over decades to evaluate eventual efficacy against cancer.

There is overwhelming epidemiological evidence causally linking HPV infection with precancerous cervical lesions and cervical cancer. The World Health Organization’s (WHO) International Agency for Research on Cancer concluded 20 years ago that there was sufficient evidence to classify many HPV types as group 1 human carcinogens (Cogliano et al. 2005). As HPV DNA detection methods have improved, studies have confirmed that oncogenic HPV DNA can be detected in almost 100% of cervical cancers and in a substantial proportion of oropharyngeal (89–95%), anal (93%), and genital cancers (63–88%) (Bosch and de Sanjose 2003; Chaturvedi 2010; Muñoz 2000). The comparison of cancer cases to controls reveals that exposure to oncogenic HPV types is associated with an enormous relative increase in the risk of developing cervical and other HPV-related cancers, consistent with a causal effect (Bosch et al. 2002; D’Souza et al. 2007; de Martel et al. 2012; Muñoz et al. 2006). The association between HPV and cervical cancer is one of the strongest ever observed for a human cancer. Though most individuals will clear their HPV infections, for many women the infections may persist for years, causing precancerous cervical intraepithelial lesions (CIN) which can eventually progress to cancer (Khan et al. 2005; Schiffman et al. 2007). Infection with an oncogenic HPV type is widely considered to be a necessary cause of cervical cancer (Franco et al. 1999; Muñoz 2000) and an important contributing cause of oropharyngeal, anal, and genital cancers (Chaturvedi 2010). This implies that all cervical cancers and a significant proportion of other HPV-associated cancers would not have occurred had the initial HPV infection been prevented. Given the solid scientific evidence supporting the role of HPV infection and high-grade lesions in the development of cervical cancer, the demonstrated efficacy against these endpoints strongly supports the assertion that the vaccine will also prevent HPV-associated cancers.

Even if we accept that the efficacy of HPV vaccines against associated cancers is undemonstrated, the opposition to vaccination on these grounds ignores the proven efficacy of HPV vaccines against other disease outcomes. The prevention of warts and precancerous lesions already provides substantial benefits to vaccinated individuals independently of eventual efficacy against cancer (Drolet et al. 2015).

In short, that clinical trials did not evaluate HPV vaccine efficacy against cervical cancer does not constitute a valid argument against HPV vaccination given the weight of the scientific evidence causally linking HPV infection with various cancers and other health outcomes. Given the long latency between onset of HPV infection and the initial neoplastic stages in cervical cancer, there has not been enough time for HPV vaccination to have made an impact on the incidence of cervical cancer; however, HPV vaccination has already had a measurable impact in reducing the incidence of precancerous lesions of the cervix (Baldur-Felskov et al. 2014; Crowe et al. 2014; Mahmud et al. 2014). Critics argue that cytology screening precludes the need for HPV vaccination due to its demonstrated effectiveness at preventing cervical cancer through detection and treatment of precancerous lesions (Rail and Lippman 2015; Tomljenovic and Shaw 2012b). The logical conclusion that preventing these lesions via vaccination will attain the same objective is lost on them.

Because HPV vaccines are prophylactic and HPV is sexually transmitted, HPV vaccination programs and recommendations generally target preadolescents before they initiate sexual activity. However, clinical trials assessed HPV vaccine efficacy in older adolescents and adults. Some have questioned whether it is warranted to recommend vaccination and implement routine vaccination programs in age groups where vaccine efficacy has not been demonstrated (Lippman et al. 2007; Nature Biotechnology 2007; Reist and Klein 2007; Thompson and Polzer 2012).

Phase III clinical trials evaluated HPV vaccine efficacy in adult populations for both ethical and practical reasons. Preadolescent populations are largely sexually inexperienced and have low rates of infection (Cubie et al. 1998). The invasive examinations required to assess efficacy against infection and precancerous lesions would not have been ethical to perform in younger populations, who would receive very little benefit from the examination. The efficacy of the vaccine would not have been assessable before many years due to the low incidence rate of infection until later adolescence and early adulthood.

Though efficacy was not assessed in preadolescents, vaccine safety and immunogenicity trials have bolstered the evidence basis for vaccine use in this age group. Comparisons of clinical trials in preadolescent, adolescent, and adult populations show that HPV vaccines have similar safety and tolerability profiles in all age groups (Block et al. 2006; Reisinger et al. 2007). Furthermore, the vaccine-induced antibody titers in preadolescents are non-inferior and potentially even superior to those observed in older participants following three doses (Block et al. 2006; Dobson et al. 2013). Experimental evidence strongly supports neutralizing antibodies as the source of vaccine-induced immunity against new HPV infections (Day et al. 2010). This suggests that the vaccine should also lead to high efficacy in younger age groups, given the high efficacy observed in adult populations with similar antibody titers.

Surveillance studies are now starting to show the impact of HPV vaccination in the first cohorts of girls vaccinated in early adolescence. A 64% reduction in HPV-6/HPV-11/HPV-16/HPV-18 infection prevalence has already been observed in girls 14–19-year-olds in the 6 years following vaccine licensure in the United States despite only moderate (51%) coverage (Markowitz et al. 2016). Similar important reductions in infection prevalence and genital warts incidence have been observed in girls in this age group in many other countries that have implemented HPV vaccination (Drolet et al. 2015).

In conclusion, the recommendation to vaccinate preadolescents was justified based on available epidemiological evidence at the time of vaccine licensure and is increasingly bolstered by emerging surveillance data of girls vaccinated in early adolescence.

Current HPV vaccines only protect against a handful of all HPV types. This limited protection has led to concerns that even if the vaccines are efficacious against infection with some HPV types, vaccination may still not reduce long-term cancer incidence because other non-vaccine HPV types may still cause cancer (Baden et al. 2007; Reist and Klein 2007; Tomljenovic and Shaw 2012b).

The first licensed HPV vaccines (Gardasil and Cervarix) were formulated to protect against HPV-16 and HPV-18. These types are responsible for the vast majority of cervical, anal, oropharyngeal, and genital cancers that are HPV positive (Backes et al. 2009; de Sanjose et al. 2010; De Vuyst et al. 2009; Kreimer et al. 2005; Li et al. 2011). Although approximately 11% of cervical cancers have traces of DNA of multiple HPV types, different attribution methods consistently estimate the proportion of cervical cancers due to HPV-16/HPV-18 to be around 70% (Vaccarella et al. 2011). There is also evidence that vaccines provide some cross protection against other related HPV types (Brown et al. 2009; Wheeler et al. 2012). A newly licensed vaccine (Gardasil 9) now protects against infection with the HPV types responsible for 90% of cervical cancers (HPV-16/HPV-18/HPV-31/HPV-33/HPV-45/HPV-52/HPV-58) (Joura et al. 2015). HPV vaccines thus give a broad protection against the HPV types causing the highest burden of morbidity and mortality. Furthermore, the precancerous lesions associated with types other than HPV-16/HPV-18 do not progress to cancer as rapidly or as frequently as those caused by the latter types (Kjaer et al. 2010; Schiffman et al. 2007) and, thus, are amenable to be detected by screening, an activity that has continued postvaccination.

Even before the vaccines were licensed, researchers were aware of a potential for HPV type replacement following vaccination (Elbasha and Galvani 2005). Type replacement constitutes an increase in the prevalence of non-vaccine HPV type infections following vaccination due to the vacating of the ecological niche occupied by vaccine HPV types. Type replacement had previously been observed in the case of the pneumococcal vaccine, which similarly only targeted a limited number of pneumococcal types (Weinberger et al. 2011). However, in the case of HPV, postvaccination surveillance studies have not observed strong marked increases in the prevalence of non-vaccine HPV types in vaccinated populations (Drolet et al. 2015). Postvaccination HPV type replacement is widely considered to be unlikely to occur for the following reasons:

1. Unlike pneumococcus, evidence does not suggest strong competitive interactions exist between HPV types which could cause type replacement (Chaturvedi et al. 2011; Thomas et al. 2000; Tota et al. 2013; Vaccarella et al. 2011; Vaccarella et al. 2013). Furthermore, this competitive interaction would have to be stronger than the cross protection induced by vaccines for type replacement to occur (Elbasha and Galvani 2005).

2. The potential for new HPV types to quickly evolve to fill the ecological niche left by HPV-16/HPV-18 is also very unlikely given the slow mutation rate of HPV (Van Doorslaer 2013).

3. Finally, non-vaccine HPV type infections have a significantly lower risk of oncogenic progression than HPV-16/HPV-18 infections (Guan et al. 2012; Khan et al. 2005; Kjaer et al. 2010). Even if some type replacement were eventually to occur, it is unlikely that this would substantively undermine the effectiveness of vaccination against HPV-associated cancers.

In conclusion, HPV vaccines protect against infections with the HPV types that cause the highest disease morbidity and mortality. The potential for competing risks from other types and type replacement are becoming increasingly less important with the advent of new multivalent vaccines targeting more HPV types that cause nearly all HPV-associated cancers. Moreover, any HPV types which could increase in prevalence do not have high oncogenic potential and thus would not substantially diminish vaccination effectiveness.

Despite the substantial evidence now supporting HPV vaccines’ safety, an association between vaccination and various adverse events remains one of the most contentious public controversies surrounding the vaccine (Franco et al. 2012). From the time of vaccine licensure, some researchers expressed the opinion that the implementation of large-scale programs was premature in light of the fact that there were no long-term data on the vaccine’s safety (Tomljenovic and Shaw 2012b). They advocated for more data on vaccine safety before integrating HPV vaccine into existing programs, generally invoking the precautionary principle as a justification. Over time, they have maintained that the link between HPV vaccines and various rare serious adverse events has not been given due attention by the scientific community (Dyer 2015; Tomljenovic and Shaw 2012a; Tomljenovic and Shaw 2012b). HPV vaccine safety is also a recurring concern for the public, and various anti-vaccine groups oppose HPV vaccines on purported safety grounds, as described by several investigators (Bingham et al. 2009; Darden et al. 2013; Hendry et al. 2013; Kata 2010; Ogilvie et al. 2010).

The claim that there were no long-term data on HPV vaccine efficacy before vaccine program implementation is debatable. Aluminum adjuvants in vaccines had been in use for some 60 years and are widely regarded as safe (Global Advisory Committee on Vaccine safety 2014; Lindblad 2004). Various randomized controlled trials (RCTs) of the HPV vaccines had for vaccine licensure been performed in thousands of young girls and women followed up to 4 years to assess vaccine safety, immunogenicity, and efficacy, and further RCT results in women, men, and children were also published in subsequent years (Block et al. 2006; Castellsague et al. 2015; Einstein et al. 2011; Future II Study Group 2007; Garland et al. 2007; Giuliano et al. 2015; Harper et al. 2006; Munoz et al. 2009; Paavonen et al. 2009; Reisinger et al. 2007; Schwarz et al. 2014; Vesikari et al. 2015; Villa et al. 2006). RCTs are the strongest source of evidence for efficacy and safety outcomes and are the gold standard for scientific health research. The strength of the evidence comes from the randomization of individuals to the HPV vaccine or the control group. The randomization ensures that HPV vaccinated and unvaccinated individuals are similar in terms of their risk factors for HPV infection, disease, and adverse events. When randomization is successful, differences in the rate of outcomes between the groups can then generally be interpreted as an effect of the vaccine. The pooling of data across Gardasil trials shows that the risk of serious adverse events was very similar between the 11,778 participants receiving Gardasil and 9680 participants receiving the placebo both in the 15 days following injection (0.5% and 0.4%, respectively) and over the entire study period (0.9% and 1.0%, respectively) (Food and Drug Administration 2006). Pooling of data across Cervarix trials similarly shows that for almost 30,000 girls and women who received the vaccine, the rate of serious adverse events was similar between participants receiving Cervarix and the control over the trial follow-up years (2.8% and 3.1%, respectively), and there was no differences in the onset of new chronic or autoimmune diseases between the vaccinated and control girls and women (Descamps et al. 2009). The most common adverse events reported in these trials were injection site pain, swelling, headache, fatigue, and fever, which were higher in the vaccine groups than in the control groups (Block et al. 2006; Future II Study Group 2007; Schiller et al. 2012). Overall, these results from large-scale RCTs indicate that HPV vaccines, while causing temporary adverse events in some individuals (pain, swelling, headache, fatigue, fever), do not increase the risk of overall serious adverse events or of chronic and autoimmune diseases.

RCTs cannot however evaluate the risk of very rare or very long-term adverse events. Various post-marketing surveillance studies have therefore been put into place in many countries in order to assess the ongoing safety of HPV vaccines. Since vaccine licensure, over 200 million doses of HPV vaccines have been distributed worldwide, providing much data to assess safety (Global Advisory Committee on Vaccine safety 2015). A first data source is the passive reporting of adverse reactions and case reports of diseases identified in vaccinated individuals. For example, the Vaccine Adverse Event Reporting System (VAERS) in the United States (US), the Canadian Adverse Events Following Immunization Surveillance System in Canada, and the Yellow Card Scheme in the United Kingdom collect reports of adverse event experienced by vaccine users. In the United States, as of 2014, 25,176 adverse event reports have been made to the VAERS for 67 million doses of Gardasil distributed (Stokley et al. 2014). The most commonly reported adverse events to passive reporting systems are injection site reactions, dizziness, syncope, nausea, and headache (van’t Klooster et al. 2011; Slade et al. 2009; Stokley et al. 2014). Case reports have been published of very rare and serious adverse events detected in vaccinated individuals, such as primary ovarian insufficiency, Guillain–Barré syndrome, anaphylaxis, venous thromboembolism, multiple sclerosis, cerebral vasculitis, complex regional pain syndrome (CRPS), and postural orthostatic tachycardia syndrome (POTS) (Brinth et al. 2015; Brinth et al. 2015; Global Advisory Committee on Vaccine safety 2015; Gruber and Shoenfeld 2015; Ojha et al. 2014; Slade et al. 2009). However, the causal interpretation of passive reporting systems and case reports is very limited because there is no comparator. Vaccinated individuals are still subject to a background rate of disease and mortality from other causes. Diseases and adverse health outcomes could coincidentally arise around the same time as vaccination from unrelated reasons. Passive reporting systems and case reports are most useful to identify outcomes that can be examined more thoroughly in larger epidemiological studies.

Independent researchers and regulatory agencies such as the World Health Organization’s Global Advisory Committee on Vaccine Safety regularly reexamine post-licensure surveillance data to determine whether vaccines can be causally linked to serious adverse events identified in case reports. The strongest evidence comes from observational cohort studies comparing outcomes in vaccinated and control populations. The comparison with a control population allows ascertaining whether the rate of disease in vaccinated individuals is substantially higher than would be expected in a demographically comparable unvaccinated population. For example, in a large cohort of nearly one million Swedish and Danish adolescent girls, vaccinated and unvaccinated girls had very similar incidence rates of venous thromboembolism (14 vs. 13 per 100,000 person years), epilepsy (51 vs. 72 per 100,000 person years), juvenile arthritis (38 vs. 37 per 100,000 person years), and numerous other autoimmune and neurological diseases (Arnheim-Dahlström et al. 2013). One French study showed a small absolute increase in the risk of Guillain-Barré syndrome in vaccinated girls (1/100000) (Agence nationale de sécurité du médicament et des produits de santé 2015), but this result was not replicated in other studies (Gee et al. 2011; Grimaldi-Bensouda et al. 2014; Slade et al. 2009). Overall, comparative studies have time and time again concluded that the incidence rate of serious adverse events in vaccinated individuals is consistent with the background rates of chronic, neurological, and autoimmune diseases and that there is very little evidence suggesting that HPV vaccination causes any of these diseases (Arnheim-Dahlström et al. 2013; Chao et al. 2012; Donegan et al. 2013; Gee et al. 2011; Global Advisory Committee on Vaccine safety 2014, 2015; Grimaldi-Bensouda et al. 2014; Scheller et al. 2014; Scheller et al. 2015).

Most of the controversy generated over HPV vaccines’ safety has resulted from selective reporting of safety data. Attacks on the vaccines’ safety generally cite only case reports/passive reporting systems while ignoring or not reporting the stronger evidence from RCTs and comparative studies. The media often seizes on case reports of rare and serious illnesses in vaccinated individuals due to their sensational and emotive nature. Despite the rarity of these diseases and the lack of evidence supporting any causal link with the HPV vaccine, these reports have strong effects on the public’s perception of the risks of vaccination. For example, a recent series of case reports of CRPS and POTS in vaccinated individuals triggered a review of the evidence by the European Medicines Agency . After a careful analysis of the case reports and epidemiological data, the agency found no evidence that the occurrence of these syndromes in vaccinated girls was different from what was expected in this age group (Pharmacovigilance Risk Assessment Committee 2015). However, in Japan the mass media and social media coverage of cases had instigated a public hysteria. In response, the Japanese Ministry of Health, Labour, and Welfare suspended the active recommendation of HPV vaccination in 2013, a decision that was politically rather than scientifically motivated. The suspension of recommendations undermined public confidence in the HPV vaccine and led to the plummeting of vaccine coverage from approximately 70% to 8% (Hanley et al. 2015; Konno et al. 2015a, b).

In conclusion, strong and consistent epidemiological evidence from both pre-licensure and post-licensure studies confirms that HPV vaccines are safe and are not causally associated with serious adverse effects. The safety of the vaccine has been continuously examined by the scientific community over the years, and no signal suggesting a causal effect of the vaccine on autoimmune or neurological diseases has emerged. Unfortunately, despite this substantial body of evidence supporting vaccine safety, fearmongering and misinformation have undermined public confidence in vaccine programs in many countries.

Some have claimed that vaccination may enhance the oncogenic progression from infection to cervical intraepithelial lesions in women who are already infected (Spinosa et al. 2011; Suba et al. 2013; Tomljenovic and Shaw 2012b; Tomljenovic and Shaw 2013). This claim is based on a post hoc sub-analysis of the FUTURE I trial of the Gardasil vaccine. In women who were already infected with and seropositive to HPV-16/HPV-18/HPV-6/HPV-11 before vaccination, there was a higher incidence rate of high-grade lesions (CIN2/3) in women who were vaccinated with Gardasil (11.1/100 person-years) than in women vaccinated with a placebo (7.7/100 person-years) (Food and Drug Administration 2006).

This observation however does not provide evidence for the vaccine enhancing the progression from infection to CIN. Firstly, the observed difference between the vaccine and placebo groups was not statistically significant. This observation is thus likely attributable to the small sample size and is consistent with the HPV vaccine having no effect on the progression rate of already established infections. Secondly, the further comparison of the two groups reveals that the women vaccinated with Gardasil already had a higher prevalence of abnormal Pap smears before vaccination than the women vaccinated with the placebo. This suggests the higher incidence rate of CIN2/3 was also in part attributable to other preexisting risk factors in women vaccinated with Gardasil rather than any effect of vaccination. Finally, the same result was not subsequently observed in the FUTURE II trial: women previously infected with and seropositive to HPV-16/HPV-18/HPV-6/HPV-11 vaccinated with Gardasil had instead a lower CIN2/3 incidence rate (6.0/100 person-years) than women vaccinated with the placebo (6.3/100 person-years) (Food and Drug Administration 2006). Further studies have confirmed that HPV vaccines do not affect the clearance and progression rate of preexisting infections (Hildesheim et al. 2016; Hildesheim et al. 2007; Syrjanen et al. 2009).

In conclusion, there is no convincing evidence that the vaccine affects the persistence and progression of preexisting HPV infections. Furthermore, this concern has limited applicability to HPV vaccination programs targeting preadolescents before sexual debut, who will be largely uninfected.

The HPV vaccines target an STI. Some parents, conservative institutions, and ethicists were initially concerned that the vaccine could cause sexual disinhibition in vaccinated preadolescents (Forster et al. 2010; McQueen 2007; Smith et al. 2008; Waller et al. 2006; Zimmerman 2006). Vaccinated adolescents might potentially increase their sexual risk behaviors and promiscuity if they perceive themselves to be protected against STIs and/or due to a normalization of sexuality at young ages. A notable example occurred in 2007–2008, when Catholic bishops in Alberta and Ontario issued statements to parents and directors of the Catholic school boards, indicating that abstinence from sexual activity was the best protection against STIs, and warning against the promotion of the message that early sexual intercourse is normative (Smith et al. 2008; Wingle 2007). Various catholic school boards subsequently voted on moral grounds not to provide the vaccine in schools (CBC News 2008). These decisions were later overturned following citizen intervention efforts (Cotter 2014; Guichon et al. 2013).

Various studies have since confirmed that preadolescents and adolescents do not increase their sexual risk behaviors after being vaccinated against HPV (Bednarczyk et al. 2012; Forster et al. 2012; Smith et al. 2015). For example, in the months following a catch-up vaccination program in England, 6% of vaccinated and 8% of unvaccinated girls 16–18 reported having initiated sexual activity since the vaccine had been offered, suggesting the receipt of the vaccine had not influenced sexual initiation rates (Forster et al. 2012). In one US study of 11–12-year-old girls eligible for vaccination, the incidence rate of diagnosis for STIs and pregnancy was 0.26/100 person-years in girls who had been vaccinated and 0.25/100 person-years in girls who had been unvaccinated in the 3 years following eligibility (Bednarczyk et al. 2012). In Canada, cohorts of girls in grade 8 eligible for school-based HPV vaccination had similar risks of pregnancy and STIs (5–6%) than cohorts of girls not eligible for vaccination (6%) during their high school years (Smith et al. 2015). The vast majority of adolescent girls still perceive safe sex practices to be important after receiving the HPV vaccine (Mullins et al. 2012).

In conclusion, the evidence suggests that exposure to HPV vaccines and to HPV vaccine programs does not change young girls’ sexual behaviors, outcomes, and attitudes.

HPV vaccines are a marketable technology developed by pharmaceutical companies. The commercial development of HPV vaccines from proof of concept, to vaccine development, to production scale-up, to clinical trials in thousands of women, and to approval was an expensive and high-risk process that took over 10 years to accomplish (Inglis et al. 2006). Therefore, pharmaceutical companies have a vested interest in the vaccines’ sale and marketability. Natural suspicion arose from the start over the influence of commercial interests over policy decisions and over the utility of HPV vaccines (Gefenaite et al. 2012; Kata 2010; Porta et al. 2008; Reist and Klein 2007; Tomljenovic and Shaw 2012b).

Vaccine opponents have argued that efficacy and safety data from HPV vaccine RCTs are suspect on the grounds that most RCTs were financed by the vaccine’s manufacturers (Lippman et al. 2014; Tomljenovic and Shaw 2012b; Tomljenovic and Shaw 2013). However, in addition to the stringent oversight imposed by regulatory approval agencies, such as the US Food and Drug Administration and the European Medicines Agency, all the clinical trials were supervised by independent data monitoring committees who reviewed safety data on an ongoing basis to ensure the ethical and safety interests of trial participants. An HPV vaccine clinical trial funded by the National Cancer Institute (NCI), a public US federal agency, also later independently confirmed the efficacy and safety results obtained in manufacturer-funded RCTs (Hildesheim et al. 2014). There is no evidence to support aspersions of scientific and ethical misconduct during clinical trials. It should also be emphasized that owing to their high cost in the tens of millions of dollars and the need for them to be conducted across multiple centers and countries, HPV vaccine trials could not be funded by any public agency or charity organization. Only large pharmaceutical companies are capable of funding such trials. The aforementioned unique example of the NCI sponsorship of an HPV vaccine trial was based on an arrangement with the manufacturer (GSK) for the NCI study site (Costa Rica) to be one of the many centers for the investigation of the candidate bivalent HPV vaccine.

Because the prevalence of HPV is high in the general population, all women (and eventually all men) were considered to be a potential market for the HPV vaccine (Nature Biotechnology 2007; Rothman and Rothman 2009). Following the licensure of Gardasil in 2006, the vaccine’s manufacturer Merck aggressively marketed its vaccine in the United States. The initial marketing placed much emphasis on the risk of cervical cancer, despite the fact that cervical cancer incidence is low (and perceived as such) in developed countries (Mah et al. 2011; Rothman and Rothman 2009). Merck’s marketing tactics included lobbying for public funding of the HPV vaccine and vaccination mandates, contributions to political campaigns and women’s health groups, educational grants to professional medical associations, and direct to consumer ads (Colgrove et al. 2010; Haber et al. 2007; Rothman and Rothman 2009). In a particularly controversial example, Merck contributed thousands of dollars to the campaign of a Texas governor, who subsequently signed an executive order to make HPV vaccination mandatory which was later revoked (Nature Biotechnology 2007). Merck eventually ceased lobbying efforts for mandatory vaccination in the United States following the negative public reaction. Nevertheless, the ensuing polemic acted as a catalyst for many of the controversies surrounding HPV vaccines’ safety, effectiveness, and utility, and considerably increased the public distrust of pharmaceutical companies and government vaccination policies. Many individuals believe that vaccines are a conspiracy foisted upon the public by pharmaceutical industries and governments for profit (Kata 2010; Madden et al. 2012). The involvement of vaccine manufacturers in the policy process exacerbated this perception.

In conclusion, commercial interests have influenced policy decisions and public perceptions of HPV vaccines. However, the value of HPV vaccines is a question that has been substantially evaluated independently by many public health experts and researchers, as discussed below.

Cervical cancer screening tests have existed for many decades, and countries with screening programs have seen substantial declines in cervical cancer incidence and mortality (Gustafsson et al. 1997; Sigurdsson 1999; Vizcaino et al. 2000). While cervical cancer remains the second most incident female cancer in developing countries, it is now only the tenth most incident cancer in developed countries, in large part due to screening (Kane et al. 2012). Papanicolaou cytology, the test used for screening, is safe and acceptable to most women. Most cervical cancers are detected in under-screened or never-screened women (Andrae et al. 2008; Kirschner et al. 2011; Leyden et al. 2005). Some have argued that wide-scale HPV vaccination is not warranted in the current epidemiological context as there is a low disease burden of cervical cancer thanks to cervical cancer screening. This argument was notably used by Finnish health authorities as a justification for not implementing HPV vaccination after licensure (Syrjänen 2010). Many fear that the focus on vaccination, whose long-term value is yet unproven, could detract from the use and improvement of cervical cancer screening, whose value has been demonstrated (Harper et al. 2010; Lippman et al. 2007; Tomljenovic and Shaw 2013). However, this argument ignores the following weaknesses of cervical cancer screening which can be counteracted through vaccination.

A single cytological screening test for cervical cancer has a relatively low sensitivity (55–90%) to detect prevalent high-grade lesions (Arbyn et al. 2008). Unlike vaccination, the success of cervical cancer screening programs is predicated on repeated testing of women over their adult lives. The necessity for repeated testing and follow-ups presents a substantial burden on health-care systems, as well as on women. For example, cervical cancer screening is estimated to annually cost the US 6.6 billion USD and the UK 208 million GBP (Brown et al. 2006; Chesson et al. 2012). For every case of cervical cancer that is detected by screening, there is an additional 50–100 women with cytological abnormalities and precancerous lesions that are discovered by screening and require proper diagnosis, treatment, and/or long-term follow-up every year (Centers for Disease Control and Prevention 1994). Despite the significant efforts deployed to ensure program quality, many cancers are still diagnosed in women for whom the screening program fails due to a false negative test, inadequate management, loss to follow-up, or interval cancer incidence (Janerich et al. 1995; Kirschner et al. 2011; Leyden et al. 2005). Moreover, screening is not effective against all cervical cancers. Screening is not very effective at preventing cervical cancers in women under 25 years (Lonnberg et al. 2012; Sasieni et al. 2009) or at preventing adenocarcinomas of the cervix, whose incidence rates have been increasing in many countries (Bulk et al. 2005; Lönnberg et al. 2015; Smith et al. 2000).

Though many advocate for increasing screening compliance and reducing program inefficiencies, it is uncertain how successful such interventions would be in reducing cervical cancer incidence and mortality. Screening coverage has stalled over the past decade in many countries despite efforts (Centers for Disease Control and Prevention 2013; Habbema et al. 2012; Machii and Saito 2011), reflecting the difficulty of reaching many marginalized women who have little or no contact with health systems. The declines in incidence and mortality rates of cervical cancer have likewise plateaued in many countries (Dickinson et al. 2012; Habbema et al. 2012; Lönnberg et al. 2015; Syrjänen 2010; Vaccarella et al. 2013), suggesting we may have nearly reached the maximal benefits of cervical cancer screening.

The focus on cervical cancer screening also ignores the other HPV-associated diseases that could be prevented through HPV vaccination. Genital warts are highly distressing and lead to non-negligible health-care costs (Chesson et al. 2012; Ostensson et al. 2015); their incidence has markedly declined in age groups targeted by HPV vaccination programs (Ali et al. 2013; Drolet et al. 2015). Oropharyngeal, anal, vulvar, vaginal, and penile cancers are also not prevented through cervical cancer screening but could be prevented through HPV vaccination. Rare put potentially fatal juvenile-onset recurrent respiratory papillomatosis might potentially also be prevented in the long-term by vaccination, as mothers may transmit their vaccine-induced HPV antibodies to their children (Shah 2014).

HPV vaccination also has indirect beneficial effects on unvaccinated individuals called herd effects. Herd effects occur because protected vaccinated individuals no longer become infected and transmit the infection to others. HPV vaccines are thus expected to reduce infection incidence in both vaccinated and unvaccinated individuals. This is not the case for cervical cancer screening, which benefits only the woman being screened. Some critics have claimed that realizing vaccine herd effects would require a high vaccination coverage to manifest (Harper et al. 2010), but this is demonstrably untrue as herd effects can immediately accrue following vaccination from reduced HPV transmission. For example, surveillance data has already shown that genital wart incidence decreased in unvaccinated heterosexual males following the implementation of female-only HPV vaccination programs in Australia (Ali et al. 2013) and that HPV vaccine type prevalence has declined in both vaccinated and unvaccinated female adolescents in the United States (Kahn et al. 2012; Markowitz et al. 2013).

Some argue that Pap tests and treatment procedures used in cervical cancer screening are much safer than vaccines and consequently that the large-scale use of vaccines with unknown risks is not ethically justifiable (Tomljenovic and Shaw 2012b). However, screening and treatment procedures do entail well-documented harms that should be weighed against potential adverse effects of vaccination. Cervical lesion treatment procedures can cause pain, bleeding, and psychological distress (O’Connor et al. 2016; Sharp et al. 2009). Cervical lesion treatment is also associated with marked subsequent increases in adverse obstetric outcomes such as preterm deliveries, miscarriages, low birth weight, and perinatal mortality (Arbyn et al. 2008; Kyrgiou et al. 2006; Kyrgiou et al. 2014). These adverse obstetric effects are not associated with HPV vaccination (Baril et al. 2015; Garland et al. 2009). Many women undergoing screening are of childbearing age, and many of them will be at risk for these adverse effects of screening over their lifetimes. HPV vaccination may substantially reduce these adverse effects by reducing the incidence of cervical lesions.

In conclusion, though cervical cancer screening has been a very effective intervention, it is likely that countries with long-standing screening programs have already reaped most of the benefits that can be achieved through screening. Primary prevention of HPV infection through vaccination presents substantial advantages, notably herd effects, and the prevention of a variety of health outcomes that cannot be prevented through cervical screening alone. The pitting of vaccination against screening is counterproductive and presents a false dichotomy, as both should be deployed as part of a comprehensive cervical cancer prevention program.

HPV vaccines are among the most expensive childhood vaccines on the market. Upon licensure, a three-dose course of Gardasil cost some 360 USD (The Lancet 2013). Over the years, critics have questioned whether public financing of HPV immunization programs is a cost-effective use of resources (Lippman et al. 2008; Porta et al. 2008; Syrjänen 2010; Thompson and Polzer 2012; Tomljenovic and Shaw 2013), given that (as per their reasoning) (1) the long-term benefits of HPV vaccination are uncertain, (2) the cost of the vaccine is high, and (3) the incidence of cervical cancer is low in most developed countries due to effective cervical cancer screening programs. Furthermore, the argument went, as the duration of vaccine efficacy is uncertain, booster shots may be required over time to maintain protection, further increasing the cost of HPV vaccine programs.

Most decision modeling analyses however agree that vaccinating preadolescent girls against HPV represents a cost-effective intervention in high-income countries using reasonable willingness-to-pay thresholds (Brisson et al. 2013; Jit et al. 2008; Konno et al. 2010; Olsen and Jepsen 2010; Seto et al. 2012). This is due largely to the sizeable gains in life years from averted cervical and other HPV-related cancers but is also due to the projected increases in quality of life and the costs saved from reduced treatment and management of vaccine-preventable cervical cancers, high-grade cervical lesions, and genital warts. In other words, providing the vaccine to preadolescent girls is generally concluded to give good value for money even at a high cost per dose and on top of existing cervical cancer screening. Public health HPV vaccination recommendations often explicitly factor in these cost-effectiveness considerations (Canadian Immunization Committee 2014; Markowitz et al. 2007). Modeling analyses do predict that vaccination of women past adolescence becomes decreasingly cost-effective (Jit et al. 2008; Kim and Goldie 2008), which supports vaccine policies targeting preadolescent girls before sexual debut. This is because many women will already have been infected by late adolescence and adulthood, reducing the cost-effectiveness of vaccination at these ages.

The high retail cost of the vaccine and lack of infrastructures for vaccine delivery in adolescents have constituted significant barriers in resource-restrained settings with competing priorities (Kane et al. 2012). HPV vaccination is predicted to be cost-effective in most low- and middle-income countries, however, when assuming a tiered vaccine cost by country according to income (Fesenfeld et al. 2013; Goldie et al. 2008; Jit et al. 2014). These countries generally have a much higher cervical cancer incidence and mortality than high-income countries and would substantially benefit from vaccination. Not surprisingly, however, cost-effectiveness analyses are highly sensitive to the vaccine price and discount rate.

Modeling analyses have consistently concluded that the vaccine protection should last at least 10–20 years in order for vaccination of preadolescents to be cost-effective, as these will constitute the years during which they will be most at risk for HPV infection (Elbasha et al. 2007; Jit et al. 2008; Kim and Goldie 2008). Though the duration of HPV vaccine protection remains uncertain, current evidence suggests that HPV vaccines should provide long-lasting immunity. The longest clinical trials which accumulated almost 10 years of follow-up data did not show any waning of vaccine efficacy against HPV-16/HPV-18 infections and associated lesions over time, which suggests that protection lasts much longer than 10 years (Ferris et al. 2014; Naud et al. 2011). Models of antibody titers predict that mean antibody titers will remain above those associated with natural infection for at least 20 years (David et al. 2009; Fraser et al. 2007). These lines of evidence suggest HPV vaccine protection should be long-lasting; there is no current indication that booster shots will be needed to maintain protection. Furthermore, in light of the natural history of HPV infection and cervical cancer, it is conceivable that the critical period for protection is during the late adolescence years, when the uterine cervix is at its most vulnerable phase with exposure of the metaplastic epithelium in the ectocervix (Schiffman et al. 2007). Vaccination in the preteen years should thus offer maximum protection even if it eventually declines.

In conclusion, economic analyses suggest that HPV vaccine programs targeting adolescent girls are a cost-effective intervention in many settings, even in countries with existing screening programs.

Because public health interventions such as vaccination are enacted to increase the overall health of the population possibly against individual preferences, they can be perceived as paternalistic and infringing upon individual self-determination (El Amin et al. 2012; Schmidt 2012). The past few decades have seen a societal shift toward the rejection of paternalism, the rise of the well-informed patient, and skepticism of science and authority (Gray 1999; Kata 2010). Many controversies hinge upon a perceived undermining of self-determination in HPV vaccination practices due to misleading communication strategies, lack of information, and coercive methods.

Some have criticized the framing of HPV vaccines by health authorities and vaccine manufacturers as anticancer vaccines addressing a public health crisis, conflating HPV with cervical cancer (Mah et al. 2011; Rail and Lippman 2015; Thompson 2013; Thompson and Polzer 2012; Tomljenovic and Shaw 2012b; Tomljenovic and Shaw 2013). Critics advance that the risk of cervical cancer and other HPV-associated cancers has been misleadingly amplified and the harms of vaccination concealed in communications in order to increase the public acceptability of HPV vaccines. Consequently, they argue that HPV vaccination practices are not ethical given that parents and children cannot give an informed consent to vaccination (Lippman et al. 2014; Rail and Lippman 2015). For example, promotional materials for HPV vaccines have emphasized cervical cancer as the second leading cause of female cancer mortality worldwide while failing to distinguish that this mortality rate is much lower in developed countries due to screening (Rothman and Rothman 2009). However, as discussed in previous sections, analyses of vaccine utility suggest that HPV vaccination does provide a substantial relative reduction in cervical cancer risk for very little harm. The framing of HPV vaccines as anticancer vaccines is also consistent with vaccine programs’ stated aims of reducing the health burden of HPV-associated cancers and not the eradication of HPV infection.

Parents still often feel they lack the necessary information to assess the harms and benefits of HPV vaccination and make an informed decision (Hendry et al. 2013). Public health agencies have over the years employed various communication strategies aimed at parents and preadolescents such as promotional materials, awareness campaigns, in-class education sessions, and consent forms (La Vincente et al. 2015; Watson et al. 2009; Wilson et al. 2012). However, many individuals distrust the information provided by authorities, either because they do not feel public health policies take into account their own individual needs or because of a general distrust of authority (Braunack-Mayer et al. 2015; Gefenaite et al. 2012). For example, parents in some countries fear that HPV vaccines are a government conspiracy to sterilize their daughters, notably in Peru where the government has historically enacted coercive sterilizations in the name of public health (Bingham et al. 2009; Bosch 2002). This creates the unfortunate situation where the public may reject the information provided in favor of other sources such as the Internet or social networks. Health-care provider recommendation of HPV vaccines can help increase vaccine uptake, especially where vaccination is not school-based (Cates et al. 2010; Gamble et al. 2010; Rahman et al. 2015). This may be because they are perceived to be responding to patient’s individual needs rather than acting as government agents obligated to enforce policy.

In some instances, the state has taken coercive action to increase HPV vaccination coverage. In the United States, bills were introduced in 23 states to make HPV vaccination mandatory for school attendance, with two states eventually enacting the mandate (Colgrove et al. 2010). These mandates for HPV vaccines were not well received by a variety of stakeholders due to the manufacturer’s excessive influence in the policy process, the non-transmissibility of HPV in the classroom, and antipathy toward governmental coercion (Charo 2007; Colgrove et al. 2010; Haber et al. 2007). Mandates for previous vaccines had been justifiable on the public health grounds of preventing the harms from the transmission of infection in the classroom. However, HPV is not transmissible in a classroom setting, which undermines a public health justification for school-entry mandates. Moreover, HPV vaccines are generally framed as anticancer vaccines, a non-transmissible disease, which emphasizes personal care over public health and an individualistic determination of the risks and benefits of vaccination (Mah et al. 2011; Thompson and Polzer 2012). Analysts have generally concluded that coercive mandates to increase HPV vaccination lack ethical justification (Opel et al. 2008; Zimmerman 2006).

In conclusion, public health arguments have had less traction in the case of HPV vaccines than for other vaccines. Instead, the discourse surrounding HPV vaccines has revolved around the self-determination of risks and benefits of vaccination. Although public health authorities have the ethical obligation of enabling an informed consent to vaccination, it is often challenging to provide this information due to increased demand for individually tailored information and mistrust of institutions.

Because more advantaged individuals are generally better able to avail themselves of health interventions, there is concern that HPV vaccination might contribute to increase health inequalities in HPV-associated diseases (Lippman et al. 2007; Polonijo and Carpiano 2013; Thompson 2013). HPV-associated cancer incidence and mortality are generally higher in ethnic minorities, those with low education, and those living in areas of low socioeconomic status (Benard et al. 2008; Braaten et al. 2005; de Vries et al. 2015; New Zealand Ministry of Health 2008; Singh et al. 2004). The inequality between countries is even sharper: nearly 90% of all cervical cancer deaths occur in developing countries, ostensibly due to low screening coverage and availability (Torre et al. 2015). Vaccine uptake is affected by the social determinants of health, and the disadvantaged individuals who would benefit most from HPV vaccination may also be the least likely to get vaccinated. The lack of resources and infrastructures in low-income countries which impede cervical cancer screening also impede the implementation of HPV vaccination programs. Of the 57 countries having implemented HPV vaccination by 2014, only a minority were low-income countries (Herrero et al. 2015).

One potential response to this issue would be vaccination strategies targeting groups at higher risk of HPV infection and HPV-associated diseases. However, previous experience with the hepatitis B vaccine showed that the targeted vaccination of high-risk groups such as intravenous drug users and men who have sex with men (MSM) did not lead to a high vaccine uptake in the United States, which led to the decision to recommend universal infant vaccination instead (Centers for Disease Control and Prevention 1991). Targeted HPV vaccination would therefore be unlikely to succeed in preventing HPV infection and reducing health inequalities. Targeted vaccination strategies also run the risk of stigmatizing the targeted population. While some have argued that universal access to vaccination does little to address the plight of marginalized social groups (Mah et al. 2011; Thompson and Polzer 2012), universal vaccination can arguably be considered to be in line with a social justice objective.

Publically funded school-based HPV vaccination programs have demonstrably been most effective at increasing overall vaccination coverage and increasing vaccination coverage in more disadvantaged groups (Hansen et al. 2015; Hughes et al. 2014; New Zealand Ministry of Health 2008; Sinka et al. 2014). Marginalized preadolescents may underutilize health services but generally have a high school attendance rate in developed countries. For example, a study in Canada showed that clinic-based delivery led to decreased vaccination coverage in girls from low socioeconomic neighborhoods compared to girls from high socioeconomic neighborhoods (34 vs. 41%), while school-based delivery led to increased vaccination coverage (83 vs. 79%) (Musto et al. 2013). Similarly, school-based delivery in New Zealand helped achieve a higher vaccination coverage in Pacific and Maori girls (78–88%) than in girls of European descent (63%) (Poole et al. 2012).

Vaccine reimbursement also substantially affects uptake. The inclusion of HPV vaccines in the Vaccines for Children reimbursement program in the United States likely contributed to increasing the vaccine coverage in adolescent girls living below the poverty level (67%) than in girls who live at or above the poverty level (55%) (Elam-Evans et al. 2014).

Universal vaccination also provides more herd effects, which can contribute to reducing health inequalities. Disadvantaged social groups with a higher cancer incidence are predicted to have larger absolute health gains from vaccination at equal coverage (Blakely et al. 2014). Even if disadvantaged individuals have a lower vaccine uptake, they should still benefit indirectly from the reduced transmission of HPV. Even if girls who eventually under-screen have a low vaccine uptake, their incidence rate of cervical cancer is likely to decrease, and absolute inequalities in the incidence rate of cervical cancer are likely to diminish postvaccination (Malagon et al. 2015). Unvaccinated girls should therefore still indirectly benefit from universal vaccination programs.

HPV vaccines could also reduce inequalities between countries, as vaccine delivery may be more feasible to implement in some contexts than routine cervical cancer screening (Tsu and Levin 2008). Programs are underway to bridge the health equity gap between high- and low-income countries. Starting in 2011, tiered vaccine prices allowed the introduction of HPV vaccines in some middle-income countries. In 2013, the GAVI alliance started supporting the introduction of HPV vaccines at lower prices in developing countries thanks to a price agreement with Merck & Co. (The Lancet 2013). Since then, GAVI has approved their support for demonstration programs in 20 countries and the national introduction of vaccination in Rwanda, Uganda, and Uzbekistan (Hanson et al. 2015). However, unlike cervical cancer screening, the benefits from HPV vaccination in low-income countries would still require decades to accrue and are unlikely to reduce inequalities in the short term.

In conclusion, questions of social justice have been and continue to be important in shaping HPV vaccination practices worldwide, reflecting the increasing attention being given to health equity in public health.

HPV vaccination was originally framed as a woman’s health issue (Mah et al. 2011). Cervical cancer has the highest incidence rate of all HPV-associated cancers and has the strongest causal link with HPV infection (de Martel et al. 2012). HPV vaccines were initially tested and approved for use in women, and most vaccination programs initially targeted girls only. However, there has always been discomfort with the idea of a gender-targeted vaccine (Prue 2016; Tjalma and van Damme 2006). Men also suffer from HPV-associated cancers and genital warts. Unlike cervical cancer, there is no screening program to prevent HPV-associated cancers in men, thus HPV vaccination represents a unique preventive intervention for these cancers. Because HPV is transmitted sexually, the vaccination of both boys and girls is also seen as a way for both genders to share the responsibility for sexual health (Luyten et al. 2014; Thompson and Polzer 2012). Nevertheless, the routine vaccination of boys was not universally adopted following the quadrivalent HPV vaccine’s FDA approval for boys in 2009 (Centers for Disease Control and Prevention 2010).