Primary prophylactic human papillomavirus vaccination programs: future perspective on global impact

Abstract

Of the 40 types of human papillomavirus that can infect the mucosal epithelium, four types can now be prevented using prophylactic vaccination. Two of these types (high-risk types 16 and 18) cause 70% of cervical cancers, a proportion of other genital cancers and a subset of head and neck cancers. The low-risk types 6 and 11 cause 90% of genital warts and the disease recurrent respiratory papillomatosis. Thus, if primary HPV vaccination programs can be implemented effectively, the potential for a reduction in global disease burden is great. This article considers the current issues and challenges in delivering primary HPV vaccination programs effectively and the likely impact of the vaccines in both the near and more distant future.

Keywords::

• cervical cancer prevention
• cervical intraepithelial neoplasia
• cervical screening
• genital warts
• HPV DNA
• human papillomavirus vaccination
• vaccine safety
• vaccine surveillance

Human papillomaviruses (HPVs) are small (55 nm), nonenveloped viruses that have circular double-stranded DNA. The circular HPV genome contains a control region, an early (E) region and late (L) region, which code for eight proteins. The six early proteins are required for virus replication and cell transformation, and the late proteins L1 and L2 are proteins of the viral capsid, against which most of the host immune response is directed [1]. L1 is the major capsid protein. HPV genotypes are distinguished by a difference in the L1 gene of at least 10% [2].

There are 40 distinct HPV genotypes that affect the genital tract and, of these, at least 13 can be categorized as high-risk types that are associated with the development of invasive cervical cancer [3]. HPV genotypes 16 and 18 were first categorized in 1995 by the International Agency for Research on Cancer (IARC) as human carcinogens. By 2009 the IARC had classified a further ten HPV types as carcinogenic to humans (31, 33, 35, 39, 45, 51, 52, 56, 58 and 59) and type 68 as probably carcinogenic [4]. Most of these types are phylogenetically related to either HPV16 (31, 33, 35, 52 and 58) or HPV 18 (39, 45 and 59). HPV types 16 and 18 are the causative agents in 70% of all cervical cancers worldwide [5,6]. HPV6 and 11 are among the low-risk (nononcogenic) HPV types, and are associated with approximately 90% of genital warts [7]. Genital HPV types are associated with a spectrum of other anogenital diseases, including certain vulval, vaginal, penile and anal cancers, and their precursors [4]. In addition, genital HPV types are associated with extragenital diseases, including some squamous cell carcinomas of the oropharynx and oral cavity [4], and recurrent respiratory papillomatosis [8].

Most women (∼80%) [9] will be infected with one or more HPV types during their lives. Comparatively less is known about the epidemiology of male genital HPV infection, with a recent multicountry cohort study suggesting a high and relatively constant incident rate across all age ranges in men [10]. The primary mode of acquisition is sexual transmission, with peak rates of HPV infection in a female population occurring in young women in the years following sexual debut [11]. There is a high probability of transmission following sexual exposure to a person with a productive HPV infection [12,13]. The virus infects basal epithelial cells, and therefore requires a breach of the surface epithelium to reach these cells and establish infection. Genital infection is thought to follow minor trauma normally associated with sex. HPV infection is not systemic (does not cause viremia), nor does it result in cell lysis or an acute local inflammatory response. HPV is adept at evading the host immune response through many mechanisms [14].

Cervical infection with HPV causes a range of pathological responses depending on the type of HPV and host factors. These range from no reaction, carriage of HPV without cytological changes, to a variety of cellular changes in the cervix. In the majority of women, these genital HPV infections (both high-risk and low-risk types) are transient and asymptomatic, clearing within 12–24 months, probably through cell-mediated immunity directed against the early proteins [14]. However, in a small proportion, infection will persist. Persistent infection with oncogenic HPV types substantially increases the risk of developing serious cervical abnormalities that may lead to cervical cancer [15]. Persistent oncogenic HPV infection can result in integration of HPV DNA into the host cell and increased production of the transforming proteins E6 and E7. The transforming properties of E6 and E7 are due to interference with the normal protective functions of tumor suppressor factors p53 and the retinoblastoma gene product (pRb), and through a variety of other interactions with cellular proteins [4,16,17]. Established cofactors that increase the chance of cervical cancer developing are smoking [18], herpes simplex virus 2 [19], chlamydia trachomatis [20], long-term oral contraceptive use [21], high parity and young age of first pregnancy [22], and immunosuppression, including that due to HIV infection [23].

Histologically, cervical abnormalities due to HPV infection have been referred to as cervical intraepithelial neoplasia (CIN), with three grades of severity, CIN1, CIN2 and CIN3/carcinoma in situ[24]. The corresponding cytological categories were mild dysplasia, moderate dysplasia and severe dysplasia/carcinoma in situ. In the late 1980s, the Bethesda system was adopted for cervical cytology, in which the main categories were low-grade squamous intraepithelial lesion (LSIL) encompassing changes thought to be due to HPV and mild dysplasia, and high-grade squamous intraepithelial lesion encompassing the former categories of moderate dysplasia, severe dysplasia and carcinoma in situ[25]. It was originally thought that there was an inevitable progression from low-grade abnormalities to high-grade abnormalities to cervical cancer. It is now recognized that LSIL cytology is a manifestation of acute HPV infection, and that most LSILs regress over time [26]. Most high-grade abnormalities also regress over time, but a woman with a persistent high-grade lesion has a higher risk of progressing to cancer. The absolute risk of cancer associated with a high-grade abnormality is difficult to determine from available observational data, but is estimated at less than 1% per year [27].

Invasive cervical cancer may arise from squamous cells at the transformation zone in the cervix between squamous and columnar epithelium (squamous cell carcinoma) or in the endocervical columnar epithelium (adenocarcinoma). Cancer of the cervix is characterized by a long screen-detectable preclinical phase of several years, with statistical modeling studies estimating that the average time from high-grade abnormalities to invasive cancer is between 10 and 15 years [28,29]. During this time, cytological changes can be detected in exfoliated cells by means of the Papanicolaou smear (Pap smear). Treatment of these abnormalities prevents the development of invasive cervical cancer and is the basis of cervical screening programs.

HPV-related disease burden

Cervical cancer is the third most common cancer in women (with an estimated 529,000 new cases per annum) and is currently estimated to cause 275,000 deaths per year [201]. The greatest burden of cervical cancer (over 85%) occurs in developing countries with little no or access to cervical screening programs, which have dramatically reduced the burden of cervical cancer in developed countries with well-functioning programs. High-risk regions are Eastern and Western Africa, Southern Africa, South-Central Asia, South America and Middle Africa [201]. Unfortunately cytology-based cervical screening needs to be conducted regularly and systematically using an organized approach, with quality systems ensuring sufficient uptake, laboratory services and continuous improvement, in order to have a substantial impact [30,31]. The resources required to implement and sustain such organized programs are beyond the reach of many countries. Developed countries consume considerable resources in the detection, treatment and follow-up of screen detected cervical abnormalities. Screening also has related costs in terms of psychosocial impacts [32–34] and potential adverse effects upon pregnancy, such as preterm delivery, among women treated for high-grade lesions [35,36].

Cervical cancer incidence rates increase sharply after 25–30 years of age in developed countries with the shape of the incidence curve highly dependent upon screening practices (peak and plateau in older ages varying between countries). In developing countries, cervical cancer has a linear relationship with age. This difference can be attributed to the screening programs in developed countries [15].

An analysis of US cancer registry data estimated that the total HPV-related cancer burden for noncervical cancers in that country is equal in size to their cervical cancer burden [37]. Other HPV associated cancers are notably rarer than cervical cancer even among screened populations (age standardized incidence vulva <2, vagina <1, penis <1 and anus <1 per 100,000 in most populations) [38]. Overall incidence estimates for oropharyngeal cancers are somewhat higher, but more difficult to interpret given that a smaller subset of these cancers are likely to be HPV related and tobacco and alcohol use are competing causes. Higher incidence rates have been observed for anal cancer in men who have sex with men [39] and HPV-related cancers in immunosuppressed populations [40]. Recurrent respiratory papillomatosis is also rare (incidence estimated at 1–4 per 100,000), but highly morbid [8]. By contrast, genital warts are very common but self-limiting, although with a considerable psychosocial burden owing to embarrassment, shame, pain and financial costs of treatment [41,42].

HPV vaccines

The eventual successful development of prophylactic HPV vaccines was contingent upon the discovery that L1 protein, when produced in a recombinant expression system, self-assembles into virus-like particles [43–45]. These virus-like particles mimic the natural virion but contain no HPV DNA, and are therefore neither infectious nor oncogenic. Two prophylactic vaccines have been developed and both have been demonstrated to produce high levels of protective antibodies when administered in a three-dose course [46,47]. The quadrivalent HPV vaccine protects against HPV6/11/16/18 and uses an aluminum adjuvant. Phase III clinical trials in over 17,000 women aged 16–26 years have shown that in women naive to relevant types at vaccination, it prevents type specific high-grade and low-grade CIN, vulval and vaginal intraepithelial lesions, and genital warts, with over 95% efficacy [48–52]. Efficacy has also been demonstrated in older women [53], and in males naive to relevant types at baseline against type specific persistent infection and external genital lesions (86 and 90% efficacy, respectively) [54]. The vaccine has also demonstrated efficacy against incident type-specific HPV16/18 related anal intraepithelial neoplasia in men who have sex with men (78%), with the US FDA consequently approving an indication for the vaccine to prevent anal cancer [202]. The bivalent vaccine protects against HPV16 and 18 and uses a novel adjuvant system AS04. A Phase III clinical trial in over 18,000 women aged 15–25 years has shown that in women naive to relevant types at vaccination it prevents type specific persistent infection, low- and high-grade CIN with over 92% efficacy [55,203]. Some cross-protection has been demonstrated in post hoc analyses for both vaccines against some HPV16 and 18 related types, although with a lesser degree of efficacy, with lower cross-neutralizing antibody titers and an unknown duration of effectiveness [55,56]. One vaccine manufacturer is currently trialing a new nonavalent vaccine, with the aim of providing a wider degree of protection against cervical cancer by providing targeted immunity against more HPV types [204].

At present there is no established immune correlate of protection against HPV infection [14]. The bivalent vaccine produced higher mean antibody titers against HPV16 and 18 than the quadrivalent HPV vaccine in a head-to-head immunogenicity trial [57]. The duration of protection provided by the vaccines is unknown but is over 7 years for the bivalent vaccine [58], 5 years for the quadrivalent vaccine [50,59,205] and 9 years for the monovalent HPV16 vaccine (a noncommercial precursor to the quadrivalent vaccine) [60]. Thus, it is uncertain as yet whether a booster dose will be required in the future. Reassuringly studies of recipients of both vaccines have demonstrated excellent anamnestic responses when given booster doses, with a rapid and marked rise in antibody titer [61,62].

The most common adverse event associated with the vaccines is local inflammation (pain, redness and swelling) at the injection site. Recipients of the bivalent vaccine were significantly more likely than those receiving the quadrivalent vaccine to report these local symptoms, fatigue or myalgia in the head-to-head trial [57]. Postvaccination safety surveillance systems across multiple countries have identified that the vaccines are safe; however, they may rarely cause anaphylaxis and the vaccination process may lead to syncope, as with other vaccines [63–65]. Other vaccine safety concerns that have been further investigated, but found not to occur at elevated rates, are thromboembolic events [206], Guillain–Barré syndrome [206] and multiple sclerosis [207]. Over 65 million doses of the quadrivalent vaccine [207] and over 20 million doses of the bivalent vaccine have been distributed [GlaxoSmithKline, Pers. Comm.]. The WHO’s Global Advisory Committee on Vaccine Safety regularly reviews safety information about the vaccines [66]. The vaccines are not recommended for use in pregnancy, although to date there is no conclusive evidence of adverse outcomes in women who have been inadvertently vaccinated when pregnant [66–68].

Issues & challenges for the delivery of HPV vaccines in primary prevention programs

Given the remarkable efficacy and safety of these vaccines in preventing cervical precancers, many countries have licensed them and incorporated them into their national immunization programs. The WHO “…recommends that routine HPV vaccination should be included in national immunization programmes, provided that: prevention of cervical cancer or other HPV-related diseases, or both, constitutes a public health priority; vaccine introduction is programmatically feasible; sustainable financing can be secured; and the cost–effectiveness of vaccination strategies in the country or region is considered” [69]. Delivering HPV vaccines effectively in primary prevention programs (which we define as delivering the vaccine primarily to those who are not yet sexually active, as opposed to including opportunistic delivery to those who may already be sexually active) has various issues and challenges, some of which are different to those faced by other routine childhood vaccination programs.

Target group

For maximum effectiveness HPV vaccines need to be provided before the targeted population is sexually active, because the vaccines do not have any therapeutic effect in those who are already infected with HPV [70]. While the median age of sexual debut differs across populations, most countries recommend primary HPV vaccination between the ages of 10 and 14 [71]. To date, primary HPV vaccination programs have been targeted at females, given that the major burden of HPV related disease driving vaccination is cervical cancer and the vaccine is expensive. In the future, as HPV vaccines become more affordable, they may be included in universal programs for both females and males. A program including males could potentially provide direct protection against HPV-related disease in males (genital warts, penile and oropharyngeal cancers and, particularly among men who have sex with men, anal cancer), as well as providing herd immunity benefits to their female partners. Such herd immunity benefits may be of particular importance, and a more cost-effective option, if female coverage cannot be feasibly increased [72].

Provision of vaccines to pre-adolescent/adolescent girls raises questions of how best to access this group. Systems for infant vaccination are well established and accepted in most countries, whereas routine vaccination of pre-adolescents and adolescents is less common, particularly in developing countries. Healthcare provision to this age group is generally less structured and often relies on self-presentation with acute illness rather than any form of routine preventive healthcare services. In many developed countries, the delivery of public health interventions, including immunization (e.g., hepatitis B, meningococcal vaccines, pertussis/diphtheria/tetanus booster), to this age group most effectively occurs through school-based services (e.g., Canada, UK and Australia) [73]. However, many countries presently lack any infrastructure or experience in school-based vaccine delivery and in some developing countries school attendance among girls is low. In addition, there are significant challenges to the delivery and documentation of a three-dose course within the school year. Studies of the efficacy of a two-dose schedule show that two doses may be sufficient for long-term efficacy, but further data are required [74,75].

A variety of approaches to vaccine delivery need to be considered in developing countries as most healthcare systems do not offer routine adolescent healthcare. School-based approaches, contracting out nongovernment organizations or utilization of the private sector are all possible strategies to reach adolescent girls [208]. PATH is leading HPV vaccine implementation projects in four countries – India, Vietnam, Uganda and Peru. The projects were designed to demonstrate which delivery strategies achieved the highest coverage and which were most feasible and acceptable. The strategies employed included school-based programs, as well as a variety of health-center and community-based approaches. Preliminary results show that high coverage of between 80 and 95% was achieved in all sites, with little difference between strategies; however, extensive education and community mobilization, as well as training of health staff and community leaders were employed in all sites. The projects showed that vaccine implementation was feasible and acceptable with key challenges similar to those seen in developed countries, such as scheduling all doses within one school year, effective training and community education, outreach to girls not attending school and the potential for adverse media coverage to impact on the vaccination program [209]. The WHO has produced a guidance document for countries considering introduction of HPV vaccines, which covers issues for consideration when preparing for introduction of HPV vaccines [210].

Policy environment, decision-making & cost

All countries face difficult decisions in relation to the allocation of healthcare resources. While immunization programs are well accepted, and for established vaccines among the most cost effective of all health interventions (being cost saving in many cases) [76], HPV vaccines are the most expensive vaccines to be developed yet. While prequalification of both vaccines by the WHO in 2009 (required to enable the purchase of vaccines by UNICEF and other UN agencies) will mean that central purchasing of the vaccines should lead to competitive pricing, until the vaccine price drops substantially HPV vaccines will remain out of reach of many countries. Cost–effectiveness analyses have demonstrated that in many developed countries with cervical screening programs the vaccines can be acceptably cost effective (in large part owing to averting the costs associated with the detection, follow-up and treatment of screen detected abnormalities) [72]. By contrast, the drivers of potential cost–effectiveness in countries without effective screening programs and with high rates of cervical cancer relate directly to the prevention of cervical cancer and lives saved [77]. However, the long delay before a reduction in cervical cancers is seen following vaccination may be a further impediment to the introduction of vaccination programs. It should also be noted that most cost–effectiveness analyses have been performed using static models that do not take herd immunity into account, and thus are likely to underestimate the potential benefits of vaccination. As a cancer preventing vaccine, there is community demand for the vaccine in many countries. To what extent decision-makers perceive and are able to respond to this demand will vary depending upon competing priorities, available resources and how healthcare decisions are made in particular countries. For example, the initial rejection of the quadrivalent HPV vaccine for national funding in Australia, owing to an unfavorable assessment of its cost–effectiveness at the price offered, led to an unprecedented political and public outcry and rapid political reassurance that the vaccine would indeed be publicly funded [78].

Education & communication with girls & parents

Ensuring that girls and their parents are aware of the vaccine and its purpose is important because three doses are required to complete the course over a 6-month period and actively returning to complete the course is contingent upon this understanding. In addition, basic knowledge of the vaccine is required to prevent a misunderstanding that it protects against all sexually transmitted infections and indeed all HPV types, and therefore to reinforce the continuing need for cervical screening. Thus, failure to convey the role and limitations of the vaccine could influence the health behavior of the recipient in the future and have adverse health outcomes.

Program & vaccine acceptability

Whether the fact that HPV is a sexually transmitted infection creates a potential barrier to the acceptability of the vaccination program appears to vary by country and population and may be a specific concern to some cultural or religious groups [79–85]. Having immigrant parents and living in a municipality with a higher percentage of voters for conservative Christian parties have been associated with lower HPV vaccine coverage in The Netherlands [83].

Coverage data and research studies from the USA and The Netherlands indicate that parents in those countries find it more acceptable to vaccinate their daughters in mid-adolescence than early adolescence [83,86–90]. Other barriers to vaccination that have been identified in the USA, where the HPV vaccine is only available free of charge to certain eligible populations through public funding, but otherwise requires reimbursement through insurance [91], include cost and safety concerns [92]. Vaccination acceptance is associated with physician endorsement, insurance coverage, perceived support of family and friends, generally positive attitudes to vaccination, recent healthcare attendance and parental and vaccinee HPV knowledge [86–89, 92–94]. An evaluation of HPV vaccination offered through a school base program in British Columbia (Canada) found that receipt of vaccination was associated with the parent’s belief that the vaccine was effective, concern about the health of their daughter and physician endorsement. Parents who had had their daughter vaccinated were also more likely to believe it would have no influence on their daughter’s sexual behavior, had generally positive vaccine attitudes and had previously had their child vaccinated. [95]. Among parents who did not have their child vaccinated the three most common reasons were concerns over safety, a wish to wait until their daughter was older and a perception that they did not have enough information to make a decision. Similar findings have been made in Australian research studies, which have found a lack of basic knowledge about HPV and a desire to know more [96–98].

Correcting organizational barriers, such as developing systems to facilitate the consent form being received by parents and returned in school programs, may also be important to obtain improved coverage rates [98]. The UK has achieved remarkably high coverage in their initial primary cohorts through school-based delivery [211]. Whether this could be owing to intensive follow-up at the local level, facilitated by the availability of identification of individuals within each area’s responsibility for vaccination, effective social marketing or the supportive community attitudes following the highly publicized death of a young woman due to cervical cancer (which prompted a noticeable increase in cervical screening attendance) is unclear [99–101]. Research from the UK has indicated that safety concerns and a perceived lack of information are also issues there [84].

Some of the barriers to vaccination that have been identified in developed countries are also major factors in developing countries, such as lack of awareness, potential sensitivities about vaccination against a sexually transmitted infection and safety concerns [102,212]. Widespread education about the vaccine and HPV should be the cornerstone of any vaccination campaign and sufficient time should be allowed for in the lead up to the implementation of a vaccination program. A rushed introduction of a program may lead to distrust by the public and media, potentially undermining the program’s success [212]. Such a situation can occur in the context of vaccine donations in low resource settings, which although given with good intentions, can divert resources from other health priorities and may not come with funds to support program implementation [103]. Severe shortages of healthcare workers, insufficient management skills and weak monitoring and information systems are known barriers to effective immunization delivery in many developing countries [104]. Adverse media coverage due to case reports of vaccine safety concerns have occurred in many developed and developing countries in the early phases of vaccination implementation and should be anticipated and prepared for by briefing key media and policy stakeholders [105,106].

Impacts of HPV vaccination

In considering the global future impact of primary HPV vaccination programs, we have grouped related impacts into two main categories: biological vaccine impacts and system impacts.

Biological vaccine impacts

Measurement issues

The reduction in HPV infection and HPV-related disease burden will be relatively complex to monitor. Outcomes of interest include short-term impacts on type-specific HPV infection rates and genital warts (months), intermediate-term outcomes, such as reductions in incident cervical lesions and juvenile-onset recurrent respiratory papillomatosis (years), and long-term outcomes, such as incidence and mortality from cervical cancer, anal cancer, other anogenital cancers and oropharyngeal cancers (decades). Each of these outcomes needs a different surveillance infrastructure for monitoring, which involves a different time frame, population sample, group of stakeholders and methodology. Added to this complexity is that each country has a different HPV vaccination program (including vaccine type, time of implementation, target population and coverage achieved), meaning that comparisons of outcomes across countries will be challenging. Such monitoring may also be relatively resource intensive, and the WHO has advised that HPV disease monitoring is not a prerequisite or an essential requirement for an HPV vaccination program [69]. Clearly it is important that these outcomes are monitored well in selected places, with timely dissemination of findings to benefit all. Postvaccination surveillance is being intensively undertaken as an extension to the vaccine trials in some of the Nordic countries [107] and developed countries who commenced HPV vaccination relatively early, such as the USA, Australia, the UK and Canada, are among those able to resource various aspects of impact assessment, which will be of benefit to all [108,109].

In late 2009, the WHO held a meeting to discuss the surveillance and coverage assessment for HPV vaccination [110]. The highlighted issues included: the importance of recording cumulative coverage data for vaccinated cohorts; the complexities in interpreting routine data from cervical screening programs; the need for robust HPV vaccination safety surveillance systems; the desirability of considering type specific HPV prevalence in young women from sentinel sites as an early indicator of vaccine impact; the priority that should be given to ensuring cervical cancer registries operate well across the globe; and the ongoing challenges in the accurate measurement of HPV type. Countries should avail themselves of the expertise and support available through the laboratories that are part of the WHO HPV Labnet, which has developed international standards for HPV testing. The monitoring of vaccination programs in developing countries is a particular challenge as few countries have well-developed cervical cytology registries and cancer registry data may be incomplete.

Expected global impacts

Modeling studies predicted a rapid fall in the incidence of type specific HPV infection with adequate vaccine coverage [12,111] followed shortly thereafter by a decline in the incidence of cervical abnormalities, and eventually cancer, which will take many years; however, the eventual impact should be substantial [112,113]. The greatest impact of the vaccine is expected to be in countries that have the highest burden of HPV-related disease. The provision of HPV vaccination alone to 70% of young adolescent girls in the poorest countries could potentially prevent the future deaths of more than 4 million women in the next decade [114]. In low-resource settings, it is estimated that vaccination alone will reduce cancer risk by 40–50% at 70% coverage [77,115], but the greatest reduction is estimated to result from vaccination followed by two lifetime HPV-DNA tests, offering estimated protection of at least 60% [115]. A study from Brazil estimated a similar benefit of over 60% reduction in cancer risk with vaccination at 70% coverage followed by three lifetime screens, reinforcing the need for a comprehensive approach to cervical cancer prevention [116]. Programs that offer both HPV vaccination to daughters and screening to their mothers may be an effective model to reduce the future cervical cancer burden in developing countries [213]. Both HPV infection and HPV-related cancers are significantly more common in HIV positive individuals [117]. Emerging trial data indicate that HPV vaccines are immunogenic and safe in immunosuppressed individuals [118,119]. As the burden of HIV infection is borne disproportionately by developing countries, HPV vaccination could potentially have a significant impact on HPV-related disease in these countries [120].

An important challenge in developing countries is how to get HPV vaccination and indeed cervical cancer control on the national agenda when competing for limited resources and other health conditions with high impact and mortality. It is clear that cervical cancer control must adopt an organized and comprehensive approach utilizing both primary and secondary prevention, and build upon the resources already available within a specific country. For example, if there is no cytology program in existence, a country may wish to consider visual inspection with acetic acid (VIA) as the initial approach. A comprehensive strategy needs to be developed and it is important to obtain appropriate information and data for further planning and monitoring the program in the initial stages [214].

Although to date there is no published evidence of an observed population level decline in HPV16/18 prevalence, there are data from Australian sexual health clinics of a decline in genital warts and from a state-based Australian Pap test Registry of a decline in high-grade cervical disease since the implementation of HPV vaccination targeted at all women aged 12–26 years in 2007. Donovan et al. have observed a 59% decline in incident genital warts diagnosis among age eligible Australian women at eight sexual health clinics around Australia. A 39% decline was also evident among heterosexual men of the same age, but not among older women or men who have sex with men [121]. Brotherton et al. have observed a decline in the incidence of histologically confirmed high-grade cervical lesions reported to the Victorian Cervical Cytology Registry among young women in Victoria (Australia) soon after the implementation of the program [122,215]. The timing of when other populations can expect to see an impact on the diagnosis of cervical lesions will be dependent upon the age of screening initiation, age at vaccination and extent of catch up programs, as well as vaccine coverage achieved [112,123]. Ultimately reducing cervical cancer incidence and mortality through vaccination in developed countries is somewhat dependent upon ensuring that vaccination reaches those girls least likely to attend cervical screening, not just those who would have been screened in the absence of vaccination. Attention to equity in vaccine provision to groups currently at the highest risk of cervical cancer should thus be a priority [124–126].

The timing of the impact of vaccination on prevalence of cervical abnormalities is particularly important, as cervical cytology screening programs will require a redesign when prevalence falls substantially. The positive predictive value of a positive screening result will fall as high-grade disease due to HPV16/18 becomes rare [127]. In this environment, it may also be difficult to maintain the performance of cytologists who will individually see fewer high-grade abnormalities. Screening at current intervals will also become less cost effective postvaccination as the number of lesions detected and treated declines. New cervical screening algorithms will most likely incorporate a primary HPV DNA test as the initial screening test, as accumulating evidence from randomized trials demonstrates that, compared with conventional cytological screening, a negative HPV DNA test has a superior and more durable negative predictive value against CIN3+ in the following years [128]. A primary HPV DNA test has the advantage that it will immediately stratify the woman into a current risk group based on the presence or absence of oncogenic HPV DNA at the time of screening, regardless of her vaccination history. A difficulty to be resolved is the subsequent method of triage of women with positive HPV DNA tests, as many of these women will have incident infection that will resolve without intervention. Potential triage strategies include cytology, progression markers and HPV genotyping [129,130].

Because present vaccines do not cover all HPV types, and therefore cannot replace screening, it is recommended in developing countries that a comprehensive approach to cervical cancer prevention is utilized that combines vaccination for young girls and screening for women over 30 years of age, perhaps only one to three times in their lifetime [114,212]. HPV testing is particularly promising in such a combined strategy owing to the high negative predictive value of testing. A trial of primary HPV testing in rural India compared with cytology or VIA found that a single lifetime HPV test significantly reduced the number of advanced cancers and deaths from cervical cancer [131]. Modeling studies that compare vaccination and combined vaccination/screening approaches also show the maximum benefit of a comprehensive approach [77,115,116].

System impacts

The other type of impact that primary HPV vaccination programs will have globally are system impacts. By this we mean that the consequences of funding and delivering HPV vaccines may have either positive or negative effects on other aspects of health, education or society. By choosing to implement HPV vaccination there will always be an opportunity cost involved, given that the resources allocated to purchasing the vaccine and providing the program could have been used elsewhere, either in health or in other areas. While an obvious concern may be that HPV vaccination programs will draw away resources from other areas, their impact may also be positive if they bring entirely new resources to the health sector and if their implementation leads to new stakeholder alliances, improved vaccination delivery, safety and coverage monitoring and cancer registries, the development of adolescent health services, normalization of school vaccination services and a prioritization of women’s health and equity.

Measurement of system impacts requires a multidisciplinary and mixed method approach, given that the impact of a program can be measured in economic terms, social terms and health terms, as well as by using ‘hard’ outcomes, such as hours or monetary costs, or ‘soft’ outcomes, such as community awareness and attitudes. Many immunization programs are systematically evaluated in the years following their implementation, usually as a one off program evaluation using data gathered from a variety of sources [132]. Ideally such evaluations are planned from the outset of the program. The long time scale from HPV vaccine implementation to biological impact may mean that system impacts also take longer to occur (e.g., the development of cancer registries). As with any new vaccine program, the immediate changes that are required may produce short-term impacts (such as changes in staffing levels or vaccine safety scares reducing confidence in vaccines generally) that may be very different to the longer-term impact of the vaccine.

Undoubtedly the major concern in relation to possible adverse impacts of HPV vaccine programs is related to their current high retail cost, at approximately US$120 per dose in developed countries [208]. Each country needs to determine what price should be paid and can be sustained for HPV vaccination, considering not only the direct costs of the vaccine, but also the costs of its implementation and whether there are more effective ways to improve health with the same budget; however, more innovative approaches are required for developing countries to obtain the best price for the vaccine and newer technologies [133]). Various models are used internationally for the financing and procurement of various vaccines, and lessons learned and strategies employed can potentially be adapted to the HPV vaccine context [134]. The GAVI alliance (a partnership of national governments, global agencies such as the WHO and UNICEF, the Bill and Melinda Gates Foundation, World Bank, nongovernment organizations and others) provides financial and technical support for countries that qualify for assistance. The GAVI alliance has prioritized support for HPV vaccines as part of its vaccine investment strategy for eligible countries (those with less than US$1000 per capita annual gross domestic product ) [208,216]. With the growing global market for HPV vaccines, and competition between manufacturers, negotiated prices are already beginning to decline [217]. As documented with hepatitis B vaccines, as the market grows and more manufacturers eventually enter the market, the price will fall [104].

Cost–effectiveness analyses are useful tools to assist in these decisions, but they have their limitations. These analyses rely on modeling of the biological impact of the vaccines integrated into the existing local cervical prevention strategy/natural history along with estimates of associated costs. Much uncertainty exists for many model parameters at present, although most models find that even at current prices primary HPV vaccination programs can be cost effective in relation to the acceptable benchmarks for other health interventions. Parameters that are most uncertain, but which have a large potential impact on modeling results, include the duration of vaccine protection, vaccine costs, discount rates, effect of herd immunity and vaccine coverage [72].

Many developed countries have been quick to embrace a new way to prevent cancer through the rapid adoption of HPV vaccines into their immunization programs. The growing size of the target population for vaccination internationally, as well the promise of second-generation prophylactic HPV vaccines in the near future, should result in a lowering of vaccine price. However, rapid adoption is by no means universal given the current high cost of the vaccines and existing highly successful screening programs in some countries [135]. Other important concerns are the inevitable impact of vaccination on the viability of cervical screening as it is currently undertaken, with a vocal concern in many countries about the inevitable disruption to already successful screening programs [136]. There is also concern that vaccinated women may be less likely to participate in any kind of cervical screening in the future [137].

To date HPV vaccines have provided a renewed focus in developed and developing countries on methods for effective adolescent/preadolescent vaccine delivery [73]. This may facilitate other health benefits to this population if systems are improved for vaccine and other health intervention delivery to this age group. There is some evidence that HPV vaccine administration does not adversely impact on coverage rates of coadministered vaccines [95,138].

Expert commentary & five-year view

There have been rapid advances in the field of cervical cancer prevention within a relatively short space of time. The widespread adoption of the HPV vaccine by many countries will result in a medium- to long-term decline in cervical abnormalities and cancer, the timing of which will depend on the age cohorts at which vaccination commenced and the age at which screening commences. Within 5 years, robust evidence on the impact of the vaccine on HPV prevalence, genital warts, cervical abnormalities and screening behavior will be available. Ongoing analysis of safety data following the receipt of tens of millions of doses worldwide will confirm the safety of the vaccines and identify any rare or idiosyncratic adverse effects of immunization. Accumulating data will also confirm or refute the vaccines’ relative safety when given inadvertently during pregnancy. Unfortunately, implementation of the vaccine has been more rapid in developed countries, in large part owing to the high cost of the vaccine, which may lead to an initial widening of the inequality in incidence and mortality between developed and developing countries that has emerged because of disparities in access to organized screening programs. A substantial fall in the price of the vaccines will be necessary for them to be more widely accessible in the developing world, which is already beginning to occur [217]. A drop in price may also allow some developed countries, particularly those whose coverage achieved among girls has been disappointing, to extend vaccination programs to males to provide herd immunity benefits, as well as primary benefits to males by preventing genital warts, penile and anal cancers and potentially some oropharyngeal cancers. In addition, remaining questions about the efficacy of a two-dose schedule, duration of protection (including whether a booster may be required) and whether the vaccine could be delivered as part of the infant immunization program may be addressed within this time period, leading to important changes to how the vaccine is delivered.

In parallel, evidence from randomized trials has been accumulating about the benefits of screening algorithms using primary HPV testing and triage by cytology or other tests, in terms of negative predictive value and improved sensitivity, that will have enormous implications for cervical screening programs and this evidence should be well developed within the next 5 years as longer follow-up data become available from these trials. However, the rapid progress in the field is also an important challenge. Policy changes to existing screening programs, or indeed implementation of new screening programs to supplement vaccination programs where there is no organized screening, will take time and require the gathering of local evidence, such as demonstration projects and cost–effectiveness studies, to inform policy. In the near future, the focus will be on accumulating such country specific evidence and this may lead to changes to screening policy in selected countries, possibly those in which large screening trials have already been conducted and where there are national HPV vaccination programs in place. More widespread changes to existing screening programs will take longer to implement at an international level, but there will be increasing pressure to implement new screening programs together with vaccination in those developing countries where cervical cancer prevention is not well established.

Finally, should a higher valency vaccine become available for use in the future, screening may no longer be necessary in those cohorts receiving such a multivalent vaccine given the additional protection afforded by the inclusion of other high-risk HPV types, assuming that such a vaccine is as effective against types 16 and 18 as the currently available HPV vaccines. Any changes to screening programs in the context of such a future vaccine would thus be transitional in nature, covering those cohorts that have either received the current bivalent or quadrivalent vaccines or that are outside the vaccine target age groups. While HPVs are not potentially eradicable in the foreseeable future, owing to the characteristics of the infection (which is usually asymptomatic, may be latent and is not treatable), both of the existing prophylactic HPV vaccines promise to greatly reduce the global burden of HPV-related disease in the near future.

Key issues

• • Bivalent and quadrivalent human papillomavirus (HPV) vaccines preventing high-risk HPV types 16 and 18 have been shown to be safe and effective in the prevention of HPV infection and cervical abnormalities.

• • The vaccines have been adopted widely and rapidly in the last few years in international immunization programs, mostly in developed countries, targeted at pre-adolescent girls.

• • The major barrier to the uptake of the vaccines by developing countries, which bear the greatest burden of HPV-related disease, is the current high cost of the vaccines, although there are some specific issues to the implementation of HPV vaccines in developing countries related to school-based delivery programs.

• • The vaccines will have substantial short-to-medium- and long-term impacts on the incidence of cervical abnormalities and cancer respectively, with predicted declines in cancer rates of more than 60% if combined with comprehensive cervical cancer control.

• • The decline in the incidence of cervical lesions will impact on the cost–effectiveness and feasibility of present cytology-based cervical screening algorithms, and will make the development of new screening approaches, probably using HPV DNA testing as a primary test, imperative.

• • Additional evidence will become available in the near future to address a variety of outstanding issues, such as number of doses required, duration of protection, potential for increasing valency of the vaccines and new algorithms for screening, all of which may have a profound impact on vaccine delivery and cervical cancer prevention.

Financial & competing interests disclosure

Both authors are investigators on an Australian Research Council Linkage Grant for which CSL Biotherapies is a partner organization. Julia ML Brotherton was an investigator on a national human papillomavirus prevalence study which received partial, equal and unrestricted funding from CSL Biotherapies and GlaxoSmithKline. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.