Presented in part: European Research Organization on Genital Infection and Neoplasia Meeting, Paris, 23–26 April 2006 (abstract S11-2).
Background. A quadrivalent (types 6, 11, 16, and 18) human papillomavirus (HPV) L1 virus-like-particle (VLP) vaccine has been shown to be 95%–100% effective in preventing cervical and genital disease related to HPV-6, -11, -16, and -18 in 16–26-year-old women naive for HPV vaccine types. Because most women in the general population are sexually active, some will have already been infected with ⩾1 HPV vaccine types at the time vaccination is offered. Here, we assessed whether such infected women are protected against disease caused by the remaining HPV vaccine types.
Methods. Two randomized, placebo-controlled trials of the quadrivalent (types 6, 11, 16, and 18) HPV vaccine enrolled 17,622 women without consideration of baseline HPV status. Among women infected with 1–3 HPV vaccine types at enrollment, efficacy against genital disease related to the HPV vaccine type or types for which subjects were naive was assessed.
Results. Vaccination was 100% effective (95% confidence interval [CI], 79%–100%) in preventing incident cervical intraepithelial neoplasia 2 or 3 or cervical adenocarcinoma in situ caused by the HPV type or types for which the women were negative at enrollment. Efficacy for preventing vulvar or vaginal HPV-related lesions was 94% (95% CI, 81%–99%).
Conclusions. Among women positive for 1–3 HPV vaccine types before vaccination, the quadrivalent HPV vaccine protected against neoplasia caused by the remaining types. These results support vaccination of the general population without prescreening.
Anogenital human papillomavirus (HPV) infection can cause invasive cervical, vaginal, and vulvar cancer; cervical, vulvar, and vaginal intraepithelial neoplasia (CIN, VIN, and VaIN, respectively); and anogenital warts [1–4]. HPV-16 and -18 cause 70% of cervical and HPV-related vulvar and vaginal cancers; >70% of cervical adenocarcinomas in situ (AIS); 50 to 60% of high-grade CIN, VIN, and VaIN (CIN2/3, VIN2/3, and VaIN2/3, respectively) cases; and 25% of low-grade CIN (CIN1) cases. HPV-6 and -11 are responsible for 90% of genital wart cases and 10% of CIN1 cases.
Prevention of persistent cervical HPV-16 and -18 infections and related CIN has been shown with monovalent or bivalent HPV virus-like-particle (VLP) vaccines [5–7]. An effective quadrivalent HPV (types 6, 11, 16, and 18) L1 VLP vaccine improves on these vaccines by increasing their public health impact. In trials conducted in >17,500 young adult women, such a vaccine was 95%–100% effective in preventing cervical, vaginal, and vulvar neoplasias and anogenital condylomata related to HPV-6, -11, -16, and -18 in women naive for the respective HPV vaccine types at enrollment [8–10].
Efficacy trials for this quadrivalent HPV vaccine did not include an HPV screening phase. More than 25% of the 16–26-year-old women enrolled in these trials had serological or polymerase chain reaction (PCR) evidence of previous or current infection with HPV-6, -11, -16, or -18. The design of the quadrivalent HPV vaccine studies stand in contrast with a recent trial of a bivalent HPV vaccine, in which HPV infection was a contraindication to enrollment [5]. Although a vaccination strategy targeting non—sexually active (HPV-naive) adolescents is intuitively advantageous, many women who are beyond sexual debut, some of whom may have already been exposed to HPV, are likely to benefit from vaccination.
Because vaccination programs are likely to target the general population of adolescent and young adult women, some women will have been previously exposed to HPV at the time of vaccination [11]. It has been unknown, however, whether women already exposed to 1 or more HPV types included in the quadrivalent HPV vaccine would still benefit from protection against disease caused by the other HPV types in the vaccine. Furthermore, the potential for vaccine-related adverse experiences in women who had already mounted anti-HPV responses to natural infection has not yet been ruled out [12, 13].
The purpose of the present report is to address these questions by assessing the prophylactic efficacy of the quadrivalent HPV vaccine in preventing CIN, VIN, VaIN, and anogenital condylomata related to HPV-6, -11, -16, and -18 in women who were either seropositive or PCR positive for at least 1 of the HPV vaccine types at enrollment.
Data sources and objective. The combined database of 2 phase 3 efficacy trials of the quadrivalent HPV vaccine was used for the present analyses. Protocols 013 (NCT00092521) and 015 (NCT00092534) (termed FUTURE I and FUTURE II, respectively) were randomized, double-blind, placebo-controlled phase 3 clinical trials designed to investigate the prophylactic efficacy of the quadrivalent (types 6, 11, 16, and 18) HPV L1 VLP vaccine (Gardasil; Merck) on HPV-6/11/16/18—related CIN, AIS, or cervical cancer (protocol 013 coprimary end point); HPV-6/11/16/18—related condyloma acuminata, VIN, VaIN, vulvar cancer, or vaginal cancer (protocol 013 coprimary end point); and HPV-16/18—related CIN2/3, AIS, or cervical cancer (protocol 015 primary end point) 8–10].
Population and study design. Between December 2001 and May 2003, 17,622 women 15–26 years old were enrolled in the 2 trials. The trials enrolled women who at day 1 reported 0–4 lifetime sex partners. Enrolled subjects with clinical evidence of genital HPV disease at day 1 were discontinued from the study before randomization. Subjects received intramuscular injections of the quadrivalent HPV vaccine or visually indistinguishable placebo at enrollment (day 1), month 2, and month 6. Each protocol was approved by the institutional review boards (ethical review committees) at participating centers, and informed consent was received from all subjects enrolled. The designs of protocols 013 and 015 and the composition of the quadrivalent HPV vaccine have been described elsewhere [9, 10].
All subjects enrolled in the study were included within the overall safety population. This population underwent full evaluation of serious adverse experiences (i.e., events that in the opinion of the investigator substantially impacted the health of the subject or that led to hospitalization). A subset of the overall safety population, termed the detailed safety population, was also asked to fill out vaccination diary cards designed to capture all generally nonserious injection-site and systemic adverse experiences, including fevers, occurring in the days after vaccination.
Clinical follow-up and laboratory testing. Examination for the presence of genital warts and vulvar and vaginal lesions was performed at enrollment (day 1), month 3 (protocol 013 only), and months 7, 12, 24, 36, and 48 (also at months 18 and 30 for protocol 013). ThinPrep (Cytyc) cytology specimens for Pap testing were collected at enrollment (day 1), month 7, and 6–12month intervals thereafter. Cytology specimens were classified using the 2001 Bethesda System [14]. Procedures for algorithm-based cytology, colposcopy, and biopsy referral have been described elsewhere [9, 10]. Biopsy material was first read for clinical management by pathologists at a central laboratory (Diagnostic Cytology Laboratories) and then read for end-point determination by a blinded panel of 4 pathologists, as described elsewhere [7–10].
Blood samples were obtained at enrollment (day 1) for anti-HPV serological testing for HPV-6, -11, -16, and -18 by competitive immunoassays (developed by Merck Research Laboratories, using technology from Luminex) [15]. Ascertainment of HPV infection involved HPV PCR analysis performed on genital swabs obtained at enrollment (day 1), month 3 (protocol 013 only), and month 7. For each subject, genital swabs were tested for the presence of HPV-6, -11, -16, and -18 DNA. For PCR analysis, swabs, biopsy samples, and thin tissue sections cut adjacent to sectionsusedforhistopathologicalanalysiswereusedtodetectHPV DNA with primers specific for HPV-6, -11, -16, or -18.
Case definition. End points for the present analyses consisted of a pathology-panel diagnosis of CIN, AIS, condyloma acuminata, VIN, or VaIN, with HPV vaccine type DNA detected in a tissue section adjacent to the section used for histological diagnosis. The method for case counting in the analysis of prophylactic efficacy among women with evidence of prior exposure or with current infection excluded cases related to the HPV type with which the woman was infected at baseline.
Statistical methods. This post hoc analyses was restricted to the subset of subjects who received at least 1 vaccination and at enrollment had evidence of current or previous infection with 1–3 of the 4 HPV vaccine types (6, 11, 16, and 18). Each woman was evaluated for subsequent development of disease due to the HPV type or types for which she was seronegative and PCR negative at enrollment. Protocol violators were included. Case counting began 30 days after the first vaccination.
A point estimate of vaccine efficacy and the 95% confidence interval (CI) were calculated on the basis of the observed split between vaccine and placebo recipients and the accrued person-time. The statistical criterion for success (P<.05) was equivalent to requiring that the lower bound of the 95% CI for vaccine efficacy exclude 0%. An exact conditional procedure was used to evaluate vaccine efficacy under the assumption that the numbers of cases in the vaccine and placebo groups were independent Poisson random variables. If a subject developed >1 end point, she was counted as a case at the date of the first end point.
The study designs of protocols 013 and 015 are compared in table 1. The protocols were similar with respect to general design, inclusion and exclusion criteria, cervical cancer screening, and diagnostic and therapeutic intervention. However, the trials differed by subject visit schedule, age range, and Pap test interval.
Subject demographics of women who were PCR positive and/or seropositive for at least 1 HPV vaccine type at enrollment as well as of the overall population are listed in table 2. Subjects with evidence of prior or ongoing infection with 1 or more HPV vaccine types at day 1 were more likely to have been pregnant, more likely to be infected with Chlamydia trachomatis, and more likely to have an abnormal Pap test result at day 1 than were the general study population. Among subjects who were positive for at least 1 HPV vaccine type at day 1, there were more subjects with a diagnosis of high-or low-grade squamous intraepithelial lesions in the quadrivalent HPV vaccine group than in the placebo group.
In the combined protocol 013 and 015 study populations, 19.8% were seropositive for HPV-6, -11, -16, and/or -18; 14.9% were PCR positive; and 26.8% were positive by either PCR or serological analysis (2368 vaccine recipients [26.9%] and 2354 placebo recipients [26.7%]). The baseline prevalence of positivity for ⩾1 HPV vaccine type was comparable between the 2 vaccination groups. Most subjects who were positive for ⩾1 HPV vaccine type were positive for exactly 1 vaccine HPV type. Among the 4 HPV vaccine types, HPV-16 positivity at day 1 was most common (HPV-16 seropositivity, 11%; HPV-16 PCR positivity, 9%; HPV-16 seropositivity and/or PCR positivity, 20%), and HPV-11 positivity was the least prevalent (HPV-11 seropositivity, 2%; HPV-11 PCR positivity, 0.7%; HPV-11 seropositivity and/or PCR positivity, 3%). Only 0.1% of the population was positive for all 4 HPV vaccine types by serological analysis and/or PCR.
table 3 presents the sizes of the subject populations that were eligible for analyses of each end point in the prophylactic efficacy analyses on the basis of HPV positivity at enrollment. For example, 1722 subjects who were positive for HPV-6, -11, and/or -16 were eligible for evaluation of HPV-18—related end points. Of these subjects, 863 were positive for HPV-6 or -11, and 1133 were positive for HPV-16 (note that some of the subjects who were positive for HPV-6 or -11 were also positive for HPV-16). The population sizes for the other HPV types are derived in a similar manner.
Baseline human papillomavirus (HPV) positivity (by polymerase chain reaction [PCR] or serological analysis) of efficacy population, by HPV-related end point.
Given that subjects who were positive for at least 1 HPV vaccine type (but <4 types) at day 1 had evidence of more frequent and more risky sexual behavior than did the general study population, it was of interest to determine whether these subjects were at higher risk for acquisition of a new infection with an HPV vaccine type than were subjects who were naive for all 4 HPV vaccine types at day 1. table 4 displays a comparison of event rates between the placebo arms of the 2 populations. As expected, subjects who were positive for at least 1 HPV vaccine type at baseline had slightly higher rates of acquisition of new HPV-related cervical, vulvar, and vaginal diseases.
Incidence of human papillomavirus (HPV)—related cervical, vulvar, or vaginal disease caused by new infections with HPV vaccine types among placebo recipients who were positive for at least 1 HPV vaccine type at day 1 and placebo recipients who were naive for all 4 HPV vaccine types at day 1.
Among women positive for HPV-6, -11, -16, or -18 at enrollment, vaccine efficacy was 100% (95% CI, 79%–100%) for the prevention of CIN2, CIN3, and AIS—lesions that represent true precursors of squamous cell and adenocarcinoma of the cervix. The observed quadrivalent vaccine efficacy for the prevention of all CIN due to the remaining HPV types in this population was 91% (95% CI, 76%–98%) (table 5) after an average of 3 years of follow-up. All 4 cases of CIN in the vaccine group were CIN1 lesions (1 lesion related to each HPV vaccine type; all 4 subjects received all 3 doses of vaccine). One subject who was PCR positive for HPV-18 at day 1 was given a diagnosis of HPV-6—related CIN1 at month 3. Another subject who was PCR positive for HPV-18 at day 1 was given a diagnosis of HPV-16—related CIN1 at month 7. For these 2 subjects, the index infection occurred before completion of the 3-dose vaccination regimen. The remaining 2 subjects received diagnoses of CIN1 lesions at month 12. One subject was seropositive for HPV-6 at baseline and was given a diagnosis of a CIN1 lesion related to HPV-18, and the other subject was PCR positive for HPV-16 at baseline and was given a diagnosis of a CIN1 lesion related to HPV-11.
Analysis of prophylactic efficacy against cervical intraepithial neoplasia (CIN) related to human papillomavirus (HPV)—, -11, -16, or -18 in a subset of subjects who were polymerase chain reaction (PCR) positive or seropositive for at least 1 HPV vaccine type at day 1.
In this same population, the efficacy of the quadrivalent vaccine in preventing external anogenital and vaginal lesions was 94% (95% CI, 81%–99%) (table 6). All 3 cases of external anogenital or vaginal lesions in quadrivalent HPV vaccine recipients were HPV-6 related, and none of these lesions were diagnosed as potential precursors of lower genital tract cancer (VIN2/3 or VaIN2/3). One subject was seropositive for HPV-18 at baseline and was diagnosed with VIN1 and condyloma at month 7. The other 2 subjects who developed cases of external genital lesions were PCR positive for HPV-16 at baseline and received a diagnosis of condyloma—1 subject at month 7 and the other at month 24. Thus, in 2 of 3 cases, the infections that resulted in the index events are likely to have occurred before the completion of the 3-dose vaccination regimen.
Analysis of prophylactic efficacy against external anogenital and vaginal lesions related to human papillomavirus (HPV)—6, -11, -16, or -18 in a subset of subjects who were polymerase chain reaction (PCR) positive or seropositive for at least 1 HPV vaccine type at day 1.
Adverse experiences were compared between the quadrivalent HPV vaccine and placebo groups stratified into 3 cohorts: seropositive and/or PCR positive for ⩾1 HPV type, seropositive regardless of PCR status, and the overall population regardless of serostatus and PCR status (table 7). For each cohort, a higher proportion of subjects given quadrivalent HPV vaccine experienced 1 or more adverse experiences than did those women given placebo, a finding predominantly attributable to injection-site reactions. Systemic adverse experiences, serious adverse experiences, and discontinuation due to adverse experiences were similar between vaccine and placebo recipients within each cohort. Systemic adverse experiences occurring in >2% (days 1–15 after vaccination) of subjects and more frequently in the vaccine group included headache (11.1% vs. 10.7%), pyrexia (5.5% vs. 4.6%), nasopharyngitis (2.6% vs. 2.3%), and nausea (2.9% vs. 2.5%). In the overall population, 4 subjects given quadrivalent vaccine and 2 subjects given placebo had 6 and 4 serious vaccine-related adverse experiences, respectively. Serious vaccine-related adverse experiences in subjects given the quadrivalent HPV vaccine included bronchospasm, gastroenteritis, injection-site movement impairment, injection-site pain, headache, and hypertension. Serious vaccine-related adverse experiences in subjects given placebo include hypersensitivity, chills, headache, and pyrexia.
The analyses provided here demonstrate that women who are infected with HPV-6, -11, -16, or -18 benefit from administration of the quadrivalent HPV (types 6, 11, 16, and 18) L1 VLP vaccine because they are protected from infections and disease caused by HPV types for which they are naive at the start of vaccination. This benefit is important because the incidence of new HPV infections among such subjects is higher than that among women who are naive for all 4 HPV vaccine types. Administration of the quadrivalent HPV vaccine to these previously exposed individuals was generally well tolerated.
The rates of infection with HPV-6, -11, -16, or -18 shortly after sexual debut are well defined. The phase 3 clinical efficacy studies for the quadrivalent HPV vaccine enrolled an ethnically and geographically diverse population of 16–26-year-old women (mean age, 20.0 years). The mean age of sexual debut was 16.7 years. This means that, on average, 27% of subjects enrolled in protocols 013 and 015 were infected with at least 1 vaccine-related HPV type within 3.3 years after sexual debut. These findings are consistent with those of recent studies that have measured the incidence of new HPV-6, -11, -16, or -18 infection in young women [16–18]. Thus, a proportion of women who have recently experienced sexual debut will continue to be naive for all 4 HPV vaccine types and will likely benefit from the quadrivalent HPV vaccine.
Although a large proportion of 13–26-year-olds will be naive for all 4 HPV types included in the quadrivalent HPV vaccine, a proportion of subjects, increasing with age, will have been infected with at least 1 HPV vaccine type. Administration of the quadrivalent HPV vaccine has not been shown to impact the course of active infections present at the initiation of vaccination. Therefore, before implementation of a catch-up vaccination program in this age range, it is important to consider whether prescreening for the presence of HPV infection is needed. The analyses presented here strongly suggest that pre-screening before vaccination of sexually experienced catch-up cohorts of young women is not advisable, because (1) infection with all 4 HPV vaccine types is very rare; (2) women who are infected with 1–3 of the 4 HPV types targeted by the quadrivalent HPV vaccine may be at high risk for acquisition of infection with the remaining HPV type(s); and (3) vaccination is highly effective in protecting these women against such incident infections.
The final consideration in institution of catch-up vaccination programs involves the cost-effectiveness of such catch-up programs relative to the routine vaccination program as well as other public health interventions. With respect to the quadrivalent HPV vaccine, a recent study using a validated population-dynamics model of HPV infection, disease, and health care costs has demonstrated that, in countries with existing cervical cancer screening programs, addition of a program of universal vaccination of 11–12-year-old girls along with catch-up vaccination in 13–24-year-old girls and women resulted in lower rates of cervical HPV disease and lower global costs compared with the addition of a program consisting solely of universal vaccination of 11–12-year-olds [19]. Similar results were seen in a recent report of the projected clinical benefits and cost-effectiveness of a bivalent (HPV-16 and -18) vaccine [20].
On the basis of the prophylactic efficacy of the quadrivalent HPV vaccine, the data presented here, and pharmacoeconomic considerations, the Advisory Committee for Immunization Practices of the Centers for Disease Control and Prevention, the group that sets vaccination policy in the United States, has stated that quadrivalent HPV vaccine programs in the United States should be implemented using a routine universal vaccination strategy in 11–12-year-old girls, supplemented by a universal catch-up vaccination strategy in 13–26-year-old girls and women. Similar recommendations have been made in other countries.
Severe Arthus-type reactions on reexposure to a given antigen have been noted with vaccines against such conditions as tetanus and hepatitis B [12, 13]. Concern of an Arthus-type adverse injection-site reaction against the quadrivalent HPV vaccine is mitigated by findings from this analysis. Statistical analysis of non—mutually exclusive populations such as those presented in this report (PCR positive and/or seropositive subjects are included in the overall population) is not ideal and, thus, is not presented. However, there was no appreciable increase in the frequency of local injection-site reactions in women who received quadrivalent HPV vaccine and had serological evidence of antibodies to an HPV vaccine type at enrollment, compared with that in the overall population.
In conclusion, women who received quadrivalent HPV vaccine were afforded high-level protection against precancerous cervical, vulvar, and vaginal lesions and genital warts related to the vaccine type or types to which they had not been previously exposed. These data provide support to the recommendation of universal catch-up immunization with the prophylactic quadrivalent HPV vaccine to prevent cervical and lower genital tract neoplasia.
Luisa L. Villa (Department of Virology, Ludwig Institute for Cancer Research, Sao Paulo, Brazil); Gonzalo Perez (National Research Center, Group Saludcoop, Bogotá, Colombia); Susanne Krüger Kjær (Department of Virus, Hormones, and Cancer, Institute of Cancer Epidemiology, Danish Cancer Society, Rigshospitalet, Copenhagen, Denmark); Jorma Paavonen (Department of Obstetrics and Gynecology, University of Helsinki, Helsinki, Finland);Matti Lehtinen (University of Tampere, School of Public Health, Tampere, Finland); Nubia Muñoz (former epidemiologist at the International Agency for Research on Cancer, Lyon, France);Kristján Sigurdsson (National Cancer Detection Clinic, Icelandic Cancer Society, Reykjavik, Iceland); Mauricio Hernandez-Avila (National Institute of Public Health, Cuernavaca, Morelos, Mexico); Ole Eric Iversen (Womens Clinic, Haukeland University Hospital, University of Bergen, Bergen, Norway); Steinar Thoresen (Cancer Registry of Norway, Montebello, Oslo, Norway); Patricia J. García (Epidemiology, STD, and HIV Unit, Universidad Peruana Cayetano Heredia, Lima, Peru); Slawomir Majewski (Department of Dermatology and Venereology, Center of Diagnostics and Treatment of Sexually Transmitted Diseases, Warsaw Medical University, Warsaw, Poland); Eng Hseon Tay (Kandang Kerbau Women's and Children's Hospital, Singapore); F. Xavier Bosch (Institut Catala d'Oncologia, IDIBELL, Barcelona, Spain); Joakim Dillner (Department of Medical Microbiology, Lund University, Malmö, Sweden); Sven-Eric Olsson (Karolinska Institute at Danderyd Hospital, Stockholm, Sweden); Kevin A. Ault (Department of Gynecology and Obstetrics, Emory University School of Medicine, Atlanta, Georgia); Darron R. Brown (Department of Medicine, Indiana University School of Medicine, Indianapolis, Indiana); Daron G. Ferris (Departments of Family Medicine and of Obstetrics and Gynecology, Medical College of Georgia, Augusta, Georgia); Laura A. Koutsky (Department of Epidemiology, University of Washington, Seattle, Washington); Robert J. Kurman (Departments of Gynecology and Obstetrics, Patholgy, and Oncology, Johns Hopkins University School of Medicine, Baltimore, Maryland); Evan R. Myers (Department of Obstetrics and Gynecology, Duke University Medical Center, Durham, North Carolina); Eliav Barr, John Boslego, Janine Bryan, Mark T. Esser, Teresa M. Hesley, Micki Nelson, Radha Railkar, Margaret James, Carlos Sattler, Frank J. Taddeo, Annemarie R. Thornton, and Scott C. Vuocolo (Merck Research Laboratories, West Point, Pennsylvania).
We thank the Data and Safety Monitoring Board (M. Boulos, J. T. Cox, F. Langmark, J. Modlin [Chair], A. Muñoz, V. Odlind, and E. Wilkinson) and the Pathology Panel (A. Ferenczy, R. Kurman, B. Ronett, and M. Stoler). We also thank all of the study participants and investigators. Finally, we thank Liwen Xi, Margaret James, Carolyn Maas, and Haiping Zhou for assistance in preparation of the manuscript.
↵Study group members are listed after the text.
Potential conflicts of interest: L.L.V., G.P., S.K.K., J.P., N.M., K.S., M.H.-A., O.E.I., S.T., P.J.G., S.M., J.D., S.-E.O., E.H.T., F.X.B., K.A.A., D.R.B., D.G.F., L.A.K., R.J.K., and E.R.M. have received honoraria from Merck Research Laboratories. These honoraria were given for consultation work and membership in the Phase III HPV Vaccine Steering and/or Registries Oversight Committees. R.J.K. is a member of the HPV Vaccine Program Pathology Panel; as such, he has been paid for his efforts in developing the Pathology Panel's standard operating procedures and for its histopathological readings of biopsy slides. G.P., S.K.K., J.P., K.S., M.H.-A., O.E.I., S.T., P.J.G., S.E.O., E.H.T., K.A.A., D.R.B., D.G.F., and L.A.K. led clinical sites that participated in the study. These investigators were compensated for all activities related to execution of the study. J.D., S.M., D.G.F., and N.M. have been given honoraria for lectureships on behalf of Merck's HPV vaccine program. D.R.B. has been paid for consultations regarding the HPV vaccine program in men. D.G.F. has been paid for consultation regarding colposcopy quality control. K.A.A. is a member of Merck's HPV Vaccine Obstetrics and Gynecology Advisory Board and, as such, receives honoraria for his consultative work.
Financial support: Merck Research Laboratories (a division of Merck & Co., Inc.).
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy