Methods

Study objectivesThe primary prespecified objective of the present analysis was to determine whether administration of quadrivalent HPV vaccine reduces the incidence of infection of ⩾6 months’ duration or cervical disease (definitions are provided below) associated with HPV-31 and -45 (originally identified as the 2 most common HPV types associated with cervical cancer worldwide after HPV-16 and -18 [20]) and with HPV-31, -33, -45, -52, and -58 (the 5 most common HPV types associated with cervical cancer worldwide after HPV-16 and -18 [20]) in an ITT population. Other end points included the incidence of infection of ⩾6 months’ duration or disease associated with nonvaccine A9 species members (31, 33, 35, 52, and 58), nonvaccine A7 species members (39, 45, and 59), and all tested nonvaccine HPV types (31, 33, 35, 39, 45, 51, 52, 56, 58, and 59)

Data sourcesData for the analysis of disease end points was derived from the combined database of 2 pivotal phase 3 randomized controlled trials of the quadrivalent HPV (type 6, 11, 16, and 18) vaccine (Merck & Co.) known as FUTURE I and FUTURE II (protocol 013 [NCT00092521] and protocol 015 [NCT00092534], respectively). Data for the analysis of infection end points was derived from protocol 012 (NCT00092482), a substudy of protocol 013. Protocol 012 was a phase 3, randomized, double-blind, placebo-controlled immunobridging study. The design of these trials has been described elsewhere [10, 11]. Description of primary analyses for protocols 012 and 013/015 can be found in the companion article by Brown et al. [21]

End-point definitionsAll end-point definitions were prespecified. Infection was defined as detection of the same HPV type in cervicovaginal/anogenital swab samples at ⩾2 consecutive visits spaced ⩾6 months apart (±1-month visit windows) or as the presence of cervical/genital disease associated with the relevant type (with type-specific HPV DNA detected in cervicovaginal or anogenital swab samples at the visit directly before or after biopsy). Data validating the use of the 6-month-infection end point are included in the online-only appendix in Brown et al. [21]. Disease was defined as diagnosis in a tissue sample of a composite end point of CIN1–3, AIS, or cervical cancer by a 4-member pathology panel with type-specific HPV DNA detected in tissue from the same lesion, as described elsewhere [10]

Clinical follow-up and laboratory testingColposcopists were trained to locate and biopsy all discrete cervical abnormalities. Biopsy samples were processed, and adjacent histological sections of each sample were first read for clinical management by pathologists at a central laboratory (Diagnostic Clinical Laboratories) who were unaware of treatment-group assignments and HPV status. Biopsy samples were fixed in formalin, and all investigators were instructed to ship on the same day that specimens were collected. Samples were processed within 24 h to insure that nucleic acids were not compromised. All polymerase chain reaction (PCR) targets were <300 bp [22]. The stability of amplification products of this size from formalin-fixed, paraffin-embedded tissue blocks has been demonstrated [23]

Statistical analysesThese analyses were conducted in all subjects who received ⩾1 injection of quadrivalent HPV vaccine or placebo and returned for follow-up, regardless of the presence of HPV infection or HPV-related disease at enrollment. Follow-up for end-point ascertainment started after day 1

Analysis plans were developed prospectively. To address the primary hypotheses with respect to infection or disease end points, a 1-sided test (α=0.025) of the null hypothesis that the vaccine efficacy (VE; defined as 100[[1-relativerisk]) is ⩽0% was conducted. The alternative hypothesis stated that the VE is >0%. A point estimate of the VE and the corresponding 2-sided 95% confidence interval (CI) were provided. Rejection of the null hypothesis (i.e., the statistical criterion for success) corresponds to a lower bound of the CI that exceeds 0%. We used an exact analysis that accounted for the amount of follow-up (i.e., person-time at risk) in the vaccine and placebo arms

Results

Protocols 013 (including the 012 substudy) and 015 enrolled 17,622 subjects combined (8810 vaccine and 8812 placebo). Of these subjects, 17,599 (99.9%) received ⩾1 dose of vaccine or placebo. The median ± SE age of subjects was 20.0 ± 2.1 years, and the range was 15–26 years. Subjects had a median lifetime number of sex partners of 2. Detailed demographics for subjects enrolled in protocols 013 and 015 have been described elsewhere [10, 11]. At study end, 85.4%, 87.5%, and 93.6% of subjects in protocols 012, 013, and 015 had completed all scheduled visits in the efficacy follow-up period

Subjects were included in the analyses presented in this article regardless of the presence of HPV infection or HPV-related disease at enrollment. On day 1, 4.6%, 5.9%, and 0.9% of subjects had cytological evidence of atypical squamous cells of undetermined significance, low-grade squamous intraepithelial lesions, and high-grade squamous intraepithelial lesions, respectively. Of these subjects with a Pap abnormality on day 1, 56.0%, 87.2%, and 93.1% were positive to 1 or more of the 14 tested HPV types, respectively. The majority of CIN2–3/AIS lesions observed in the placebo arm were associated with HPV-16–related HPV types (A9 species). Of these CIN2–3/AIS lesions, 29.5% were associated with HPV-31, -33, -35, -52, and/or -58 with no coinfection with vaccine HPV types, and 13.7% were associated with HPV-31, -33, -35, -52, and/or -58 with coinfection with vaccine HPV types. In the vaccine arm, lesions associated with a mix of vaccine and nonvaccine HPV types were rare, due to high VE for HPV-6, -11, -16, and -18

In the combined population of protocols 013 and 015, a total of 32.8% of enrolled subjects were DNA positive to ⩾1 of 14 HPV types tested on day 1 (table 1). HPV-16 was the most common HPV type, followed by 56, 51, and 52 (table 1)

Subjects were followed for an average of 3.6 years after receipt of dose 1. A total of 3459 subjects (1732 in the vaccine arm and 1727 in the placebo arm) were included in the ITT analysis of infection from protocol 012. This population included both HPV-naive women and women with preexisting HPV infection and HPV-related disease at enrollment. Administration of quadrivalent HPV vaccine reduced the combined incidence of infection (table 2) with HPV-31/45 by 31.6% (95% CI, 15.4% to 44.7%) and with HPV-31/33/45/52/58 by 17.7% (95% CI, 5.1% to 28.7%). To address any potential ascertainment bias resulting from the higher frequency of colposcopy, biopsy, and definitive therapy among placebo recipients or from the inclusion of cervical biopsy samples in the analysis of infection (see Methods), we did supportive analyses whereby infection was restricted to detection of HPV DNA in cervicovaginal/anogenital swab samples only. In these analyses, the efficacy for HPV-31/33/45/52/58 infection was 16.9% (95% CI, 3.9% to 28.2%). Individual HPV types with statistically significant reductions in infection for the vaccine arm compared with the placebo arm included HPV-31 (33.6% [95% CI, 14.6% to 48.5%]) and HPV-59 (24.6% [95% CI, 1.9% to 42.2%]). Additional analyses of efficacy against infection of ⩾6 months’ duration were also conducted. The definition of infection was limited to detection of HPV DNA in cervicovaginal swab samples (primary analyses of infection included biopsy PCR data). Data from the swab-only infection analyses were consistent with the reported results, because the majority of cases of infection of ⩾6 months’ duration were based solely on the detection of HPV in swab samples. For example, efficacy against infection with HPV-31/33/45/52/58 of ⩾6 months’ duration determined using swab data only was 16.9% (95% CI, 3.9% to 28.2%), compared with the 17.7% (95% CI, 5.1% to 28.7%) reported above (cases were counted after day 30 in the swab-only analysis)

A total of 17,160 subjects (8562 in the vaccine arm and 8598 in the placebo arm) were included in the analysis of disease from protocols 013 and 015 (tables 3 and 4). Administration of quadrivalent HPV vaccine reduced the incidence of HPV-31/45–related CIN1–3/AIS by 22.2% (95% CI, 4.4% to 36.7%) (table 3). Efficacy against HPV-31/45– and HPV-31/33/45/52/58–related CIN1 was 24.6% (95% CI, 4.1% to 40.8%) and 23.5% (95% CI, 10.8% to 34.4%), respectively. Efficacy against the combined incidence of HPV-31/33/45/52/58–related CIN1–3/AIS was 18.8% (95% CI, 7.4% to 28.9%). Although a modest reduction in HPV-31/33/45/52/58–related CIN2 or worse was observed, the estimated reduction was not statistically significant

Combined efficacy against CIN1–3/AIS related to the 10 tested nonvaccine HPV types (31, 33, 35, 39, 45, 51, 52, 56, 58, and 59) collectively was 15.1% (95% CI, 6.0% to 23.4%) (table 4). Efficacy against CIN2 or worse related to these 10 types combined was 13.2% (95% CI, −2.0% to 26.0%) (data not shown). VE against the combined incidence of CIN1–3/AIS related to nonvaccine A9 species HPV type (31, 33, 35, 52, and 58) and nonvaccine A7 species HPV types (39, 45, and 59) was 19.2% (95% CI, 7.9% to 29.1%) and 14.7% (95% CI, −4.4% to 30.4%), respectively. Individual HPV types with significant observed efficacy against CIN1–3/AIS included HPV-31 (26.0% [95% CI, 6.7% to 41.4%]), HPV-58 (28.1% [95% CI, 5.3% to 45.6%]), and HPV-59 (37.6% [95% CI, 6.0% to 59.1%]). Other HPV types displayed positive effects, but differences were not statistically significant

The contribution of HPV-31 cross-protection to composite HPV type cross-protection (i.e., against all tested nonvaccine HPV types and against the 2 other nonvaccine type composite groupings [HPV-31/33/35/52/58 and HPV-31/33/45/52/58]) is shown in table 5. Significant efficacy against CIN1–3/AIS was observed for each of the 3 nonvaccine HPV type composites when HPV-31 was removed

Discussion

We conducted an ITT analysis and estimated the cross-protective vaccine impact for 10 nonvaccine HPV types (31, 33, 35, 39, 45, 51, 52, 56, 58, and 59) in phase 3 efficacy studies of the quadrivalent HPV vaccine. The analysis of cross-protective efficacy against cervical disease included all women in the FUTURE I and FUTURE II trials who received ⩾1 dose of vaccine or placebo. The analysis of cross-protective efficacy against infection included all women entering protocol 012 (a substudy of protocol 013 [FUTURE I]) who received ⩾1 dose of vaccine or placebo. It should be noted that, as a result of the high efficacy of the vaccine seen in FUTURE I and II, the independent data and safety monitoring board for these studies recommended ending follow-up early. The placebo arm was offered the potential benefits of vaccination as quickly as possible

Our results demonstrate statistically significant cross-protective reductions in infection and composite CIN (mostly due to reductions in CIN1) end points among a population consisting of both HPV-naive women and women with preexisting HPV infection and HPV-related disease at enrollment. Although a modest reduction in HPV-31/33/45/52/58–related CIN2 or worse was observed, the estimated reduction was not statistically significant. Previous investigations of HPV vaccine cross-protection did not report disease end points or ITT analyses, which are critical to estimating the potential public health benefits of cross-protection [24]

Within specific geographic regions, baseline prevalence and therefore cumulative HPV exposure will naturally vary among populations. This variance will be related to both sexual behaviors and to the individual HPV genotypes initially established, given unique population genetics and exposure frequencies (i.e., founder effects) [4, 25, 26]. A significant proportion (32.8%) of the FUTURE I and II trial participants were infected at the time of study enrollment with 1 or more of the genital HPV types under evaluation for vaccine cross-protection. Despite prevalent oncogenic HPV infections with both vaccine and nonvaccine HPV types, cross-protective VE against infection was observed for both composite end points as well as the individual HPV types 31 and 59. A corresponding reduction in disease, principally attributable to reductions in CIN1, was observed for the composite end point of HPV-31/33/45/52/58–related disease. A statistically significant reduction in disease due to specific individual HPV types was also observed for HPV-31/58/59

The interpretation of this study is accompanied by some limitations. Although cross-protective efficacy was detected through 3.6 years in this study, uncertainties remain regarding the durability of HPV vaccine immunity. It is possible that differences in cross-protective immunity against nonvaccine HPV types and type-specific immunity against vaccine HPV types may emerge if overall vaccine immunity wanes. In addition, our study was not powered to measure reductions in CIN2–3 due to nonvaccine HPV types in the ITT population, and a significant number of CIN2–3 end points included coinfections with HPV vaccine types (i.e., 16 and 18). Additionally, assessment of disease end points may have been confounded by reductions in colposcopic biopsy due to reductions in referrals for disease related to HPV-6/11/16/18 infection in the vaccine arm. However, assessment of HPV vaccine cross-protection for HPV infections (via swab samples only) was not affected by ascertainment bias, because samples were collected for both vaccine and placebo recipients at each indicated study visit (efficacy against infection with HPV-31/33/45/52/58 as determined using swab samples only was 16.9% [95% CI, 3.9% to 28.2%]). Despite these limitations, the reductions in type-specific infections mirror the reductions in CIN1 and provide evidence of a partial cross-protective benefit for the quadrivalent HPV vaccine

Because coinfection with genital HPVs is common, the benefit of any cross-protection against nonvaccine oncogenic HPV types will not be fully additive. This is highlighted by analyzing data from the placebo arm. In the placebo arm, there were 273 cases of CIN2–3/AIS containing HPV-16/18 DNA and 246 containing HPV-31/33/52/58 DNA. If the benefit of cross-protection were fully additive, we would expect 519 cases (273 plus 246) of HPV-16/18/31/33/52/58–related CIN2–3/AIS. However, because cross-protective efficacy is not fully additive, there were actually 436 cases of CIN2–3/AIS containing HPV-16/18/31/33/52/58 DNA. This indicates that roughly 19% (83/436) of cases of CIN2 or worse containing HPV-16/18/31/33/52/58 DNA were coinfected

The potential benefit of cross-protection is relevant to 2 separate disease risk compartments. First, in women who become infected with nonvaccine types in the absence of HPV-6/11/16/18 infection, disease reductions due to cross-protective effects of the quadrivalent vaccine on CIN of any grade might provide an incremental benefit by reducing referrals for repeat Pap testing, colposcopic procedures, or treatment. Second, an additional benefit of the quadrivalent HPV vaccine may also be provided to women at risk for coinfection with both vaccine and nonvaccine HPV types

Because overall and type-specific reductions in HPV-related disease will be evaluated over time, the current study presents an opportunity to discuss the potential relevance of existing competing disease risks, including potential niche replacement and unmasking, as observed in other vaccine settings [27–30]. Unmasking of disease risk refers to when coinfections of related etiological agents are initially unassigned due to sampling or detection bias. Unmasking due to nonvaccine HPV type–specific coinfections represents a potentially predictable competing disease risk, at least in a subset of at-risk women [29, 30]. Among some women with partners who are infected with both vaccine and nonvaccine HPV types, protection against vaccine HPV types may unmask risks associated with infection by nonvaccine oncogenic HPV type(s), and these infections may result in CIN2–3/AIS (or cancer). An intervention that reduces the risk of acquisition of infection with nonvaccine HPV types among vaccinated subjects will reduce this unmasking phenomenon—thereby further reducing the vaccinated subject’s overall risk of developing CIN2–3/AIS

The term “niche replacement” has been used to reflect the emergence of unanticipated disease risks. Changes in a biological niche can presumably alter competitive interactions among disease-related agents or can result in the emergence of more virulent or pathogenic “strains.” Niche replacement has been causally associated with some vaccines, such as 7-valent pneumococcal conjugate vaccine [31]. In contrast to unmasking of disease risks, niche replacement due to alterations in the overall fitness or carcinogenicity of nonvaccine HPV types is not supported by evolutionary data, although this presumption must be borne out through long-term follow-up. Bacterial agents with large genomes and viruses with high mutation rates can readily drive genetic adaptations; however, HPVs demonstrate little to no capacity for mutation (i.e., as few as 1 mutation or single-nucleotide substitution in 300 bp over several thousands of years) [32]

Anticipating niche replacement due to competitive interactions among HPV genotypes is a potentially more difficult task, given that natural history studies of type-specific HPV infections have presented conflicting observations. For instance, incident HPV infection after an HPV-16 infection, type-specific HPV persistence/clearance, and coinfection have in some studies demonstrated near independence [33,34]. Examples of contrasting observations include (1) a decreasing trend in the risk estimates for CIN2 or worse among women infected with HPV-16 and either 0, 1, or multiple noncarcinogenic HPV types [35]; (2) antagonism between the noncarcinogenic HPV types 6 or 11 and HPV-16 resulting in a reduced risk for both CIN and invasive cervical cancer [36, 37]; and (3) increased odds of acquiring a subsequent HPV-58 infection after an incident HPV-16 or -18 infection compared with individuals not infected with HPV-16 or -18 [38]. The potential for disease replacement driven by ill-defined existing or emerging competitive interactions among HPV genotypes after HPV vaccination remains to be determined

In long-term postmarketing surveillance programs, consideration of the above issues will be important to properly characterize any changes in type-specific HPV prevalence and disease incidence [39], particularly among cervical precancers (i.e., CIN2–3). Baseline and longitudinal measurements that compare type-specific HPV prevalence observed in women diagnosed with CIN1–3 with that observed among women with asymptomatic HPV infections (i.e., normal cytology) will be required to distinguish potential HPV type replacement from expected unmasking of competing disease risks. It will also be important for long-term evaluations of HPV vaccines to be conducted in true population-based settings and to account for any complementary modifications implemented within cervical screening programs. Even with highly effective interventions, the potential for overestimation of effects exists if replacement or unmasking is observed

In summary, administration of quadrivalent HPV vaccine to a partially infected population reduced the incidence of infection of ⩾6 months’ duration and CIN1 associated with additional nonvaccine oncogenic HPV types. Although a modest reduction in HPV-31/33/45/52/58–related CIN2 or worse was observed, the estimated reduction was not statistically significant. These results complement the vaccine’s high prophylactic efficacy against disease associated with HPV-6/11/16/18 infection. Long-term monitoring of vaccinated populations will be needed to more fully ascertain the population-based impact and public health significance of these findings. Looking toward the future, second-generation HPV vaccines targeting a broader spectrum of oncogenic HPV types may become available within the next decade, expanding the benefits of HPV prophylactic vaccines beyond the current specific and cross-protective vaccine immunity

Author affiliationsDepartment of Molecular Genetics and Microbiology and Department of Obstetrics and Gynecology, University of New Mexico, Albuquerque NM (C.M.W.); Department of Virus, Hormones, and Cancer, Institute of Cancer Epidemiology, Danish Cancer Society/Rigshospitalet, Copenhagen, Denmark (S.K.K.); National Cancer Detection Clinic, Reykjavik, Iceland (K.S.); Department of Clinical Medicine, University of Bergen and Department of Obstetrics and Gynecology, Haukeland University Hospital, Bergen, Norway (O.-E.I.); Institute of Public Health, Cuernavaca, Morelos, Mexico (M.H.-A.); National Research Center, Group Saludcoop, Bogotá, Colombia (G.P.); Department of Medicine, Indiana University School of Medicine, Indianapolis (D.R.B.); Department of Epidemiology, University of Washington, Seattle (L.A.K.); KK Women’s and Children’s Hospital, Singapore (E.H.T.); Epidemiology HIV and STD Unit, Universidad Peruana Cayetano Heredia, Lima, Peru (P.G.); Department of Gynecology and Obstetrics, Emory University School of Medicine, Atlanta, Gerogia (K.A.A.); Microbiology and Infectious Diseases Department, Royal Women’s Hospital and Department of Obstetrics and Gynecology, University of Melbourne, Melbourne, Victoria, Australia (S.M.G.); Department of Gynecology and Obstetrics, Medical University of Vienna, Vienna, Austria (S.L. and E.A.J.); Karolinska Institute at Danderyd Hospital, Stockholm, Sweden (S.-E.O.); Department of Obstetrics and Gynecology, University of Hong Kong, Hong Kong Special Administrative Region, China (G.W.K.T.); Department of Family Medicine and Obstetrics and Gynecology, Medical College of Georgia, Augusta, Georgia (D.G.F.); Department of Obstetrics and Gynecology, University Central Hospital, Helsinki, Finland (J.P.); Direction Risques Biologiques, Environnementaux et Occupationnels, Institut National de Santé Publique du Québec, Montréal, Canada (M.S.); Institut Catala d’Oncologia, IDIBELL, Barcelona, Spain (F.X.B.); Department of Medical Microbiology, Lund University, Lund, Sweden (J.D.); Departments of Gynecology and Obstetrics, Pathology, and Oncology, Johns Hopkins University School of Medicine, Baltimore, Maryland (R.J.K.); Department of Dermatology and Venereology, Center of Diagnostics and Treatment of Sexually Transmitted Diseases, Warsaw Medical University, Warsaw, Poland (S.M.); National Institute of Cancer, Bogotá, Colombia (N.M.); Department of Obstetrics and Gynecology, Duke University Medical Center, Durham, North Carolina (E.R.M.); Department of Virology, Ludwig Institute for Cancer Research, Sao Paulo, Brazil (L.L.V.); Merck Research Laboratories, West Point, Pennsylvania (F.J.T., J.B., C.R., A.T., L.C.L., K.E.D.G., M.J., S.V., T.M.H., and E.B.)

Potential conflicts of interestN.M. has received lecture fees, advisory board fees, and consultancy fees from Merck and Sanofi Pasteur MSD. S.-E.O. has received lecture fees from Merck. M.H.-A. has received lecture fees and grant support from Merck. O.-E.I. has received lecture fees from Merck and GlaxoSmithKline (GSK). C.M.W. has received funding through her institution to conduct HPV vaccine studies for GSK. K.A.A. has received consultancy and advisory board fees from Merck and has received funding through his institution to conduct HPV vaccine studies for Merck and GSK and nonvaccine clinical trials for Gen-Probe. F.X.B. has received lecture fees from Merck and GSK and has received funding through his institution to conduct HPV vaccine studies for GSK. J.P. has received consultancy fees, advisory board fees, and lecture fees from Merck. J.D. has received consultancy fees, lecture fees, and research grants from Merck and Sanofi Pasteur MSD. S.L. has received lecture fees from Merck and Sanofi Pasteur MSD. E.A.J. has received lecture fees from Merck, Sanofi Pasteur MSD, and GSK. S.K.K. has received consultancy fees and funding through her institution to conduct HPV vaccine studies for Sanofi Pasteur MSD and Digene. S.M.G. has received advisory board fees and grant support from Commonwealth Serum Laboratories and GSK and lecture fees from Merck. D.G.F. has received consultancy fees and funding through his institution to conduct HPV vaccine studies for GSK and has received lecture fees and consultancy fees from Merck. K.S. has received consultancy fees from Merck. S.M. has received lecture fees and advisory board fees from Merck. G.P. has received lecture fees and consultancy fees from Merck and Sanofi Pasteur MSD. D.R.B. has received lecture fees, advisory board fees, and intellectual property fees from Merck. M.S. has received lecture fees and grant support from Merck. Additionally, S.-E.O., C.M.W., M.H.-A., L.L.V., O.-E.I., G.W.K.T., F.X.B., J.P., J.D., E.H.T., S.L., E.A.J., S.K.K., G.P., D.G.F., K.S., M.S., L.A.K., and D.R.B. have received funding through their institutions to conduct HPV vaccine studies for Merck. F.J.T., C.R., A.T., J.B., L.C.L., K.E.D.G., S.V., M.J., T.M.H., and E.B. are employees of Merck and potentially own stock and/or stock options in the company