HPV Vaccine against Anal HPV Infection and Anal Intraepithelial Neoplasia
Download the PDF here
N Engl J Med Oct 27 2011
Joel M. Palefsky, M.D., Anna R. Giuliano, Ph.D., Stephen Goldstone, M.D., Edson D. Moreira, Jr., M.D., Carlos Aranda, M.D., Heiko Jessen, M.D., Richard Hillman, M.D., Daron Ferris, M.D., Francois Coutlee, M.D., Mark H. Stoler, M.D., J. Brooke Marshall, Ph.D., David Radley, M.S., Scott Vuocolo, Ph.D., Richard M. Haupt, M.D., M.P.H., Dalya Guris, M.D., and Elizabeth I.O. Garner, M.D., M.P.H.
The rate of anal cancer is increasing among both women and men, particularly men who have sex with men. Caused by infection with human papillomavirus (HPV), primarily HPV type 16 or 18, anal cancer is preceded by high-grade anal intraepithelial neoplasia (grade 2 or 3). We studied the safety and efficacy of quadrivalent HPV vaccine (qHPV) against anal intraepithelial neoplasia associated with HPV-6, 11, 16, or 18 infection in men who have sex with men.
In a substudy of a larger double-blind study, we randomly assigned 602 healthy men who have sex with men, 16 to 26 years of age, to receive either qHPV or placebo. The primary efficacy objective was prevention of anal intraepithelial neoplasia or anal cancer related to infection with HPV-6, 11, 16, or 18. Efficacy analyses were performed in intention-to-treat and per-protocol efficacy populations. The rates of adverse events were documented.
Efficacy of the qHPV vaccine against anal intraepithelial neoplasia associated with HPV-6, 11, 16, or 18 was 50.3% (95% confidence interval [CI], 25.7 to 67.2) in the intention-to-treat population and 77.5% (95% CI, 39.6 to 93.3) in the per-protocol efficacy population; the corresponding efficacies against anal intraepithelial neoplasia associated with HPV of any type were 25.7% (95% CI, -1.1 to 45.6) and 54.9% (95% CI, 8.4 to 79.1), respectively. Rates of anal intraepithelial neoplasia per 100 person-years were 17.5 in the placebo group and 13.0 in the vaccine group in the intention-to-treat population and 8.9 in the placebo group and 4.0 in the vaccine group in the per-protocol efficacy population. The rate of grade 2 or 3 anal intraepithelial neoplasia related to infection with HPV-6, 11, 16, or 18 was reduced by 54.2% (95% CI, 18.0 to 75.3) in the intention-to-treat population and by 74.9% (95% CI, 8.8 to 95.4) in the per-protocol efficacy population. The corresponding risks of persistent anal infection with HPV-6, 11, 16, or 18 were reduced by 59.4% (95% CI, 43.0 to 71.4) and 94.9% (95% CI, 80.4 to 99.4), respectively. No vaccine-related serious adverse events were reported.
Use of the qHPV vaccine reduced the rates of anal intraepithelial neoplasia, including of grade 2 or 3, among men who have sex with men. The vaccine had a favorable safety profile and may help to reduce the risk of anal cancer. (Funded by Merck and the National Institutes of Health; ClinicalTrials.gov number, NCT00090285.)
Our study shows efficacy of the qHPV vaccine against HPV-6, 11, 16, or 18–related anal intraepithelial neoplasia of grade 1 (including condyloma) or grade 2 or 3, against persistent anal infection with each of the four HPV strains, and against anal detection at any time of DNA of each of the four HPV types, in both the per-protocol efficacy population and the intention-to-treat population. The proportion of participants who reported serious adverse events or who discontinued the study owing to an adverse event was relatively low and was similar in the two groups. Lower rates of adverse events were observed in this study, in both groups, than in earlier studies of female participants — particularly regarding injection-site–related and systemic adverse events.10,11
Strengths of this study include the study design, as well as inclusion of participants from several countries, resulting in a diverse study population. Limitations include the narrow range of ages of the participants and the relatively short follow-up time. The study participants had limited sexual activity (a maximum of five lifetime sexual partners) as compared with many boys and men who have sex with men of similar age or older,15 and since the qHPV vaccine is preventive, the results may not be generalizable to boys and men in the general population of similar ages to those of our study population. Among men who have sex with men who have not yet initiated sexual activity, vaccination would most likely result in levels of efficacy similar to those in our per-protocol efficacy population.
HPV vaccination in men who have sex with men presents special challenges. Efficacy would be optimal if vaccination occurred before the initiation of sexual activity, but few boys identify themselves to parents or physicians as men who have sex with men by this time.16 Programs designed to target persons for vaccination on the basis of sexual orientation at a time when they have had limited prior sexual exposure would probably fail. Furthermore, the “herd immunity” that may result from vaccinating only girls and women would not fully benefit men who have sex with men, since these men may become infected with HPV through sexual contact with girls, women, boys, or men. Consistent with this, the rate of genital warts declined among heterosexual men, but not among men who have sex with men, in a setting with high levels of vaccination of girls and women.17
Although our study only included men who have sex with men, our data suggest potential benefits of vaccination for women and heterosexual men, beyond the already demonstrated protection against cervical and vulvovaginal disease and external genital condyloma. Anal HPV infection, anal intraepithelial neoplasia, and anal cancer have been shown to occur in women and heterosexual men.18-22 Given the biologic similarity between anal cancer in men and women, including the high proportion of anal-cancer cases associated with HPV-16 or 18 infection, we would expect the qHPV vaccine to protect against anal intraepithelial neoplasia in the female and heterosexual male populations to a degree similar to that among men who have sex with men.
Our study suggests that qHPV vaccination could be a tool for preventing anal HPV-related disease, potentially even cancer. There were no cases of anal cancer in this young population, as we expected. However, just as the prevention of cervical intraepithelial neoplasia of grade 2 or 3 is expected to reduce the risk of cervical cancer in vaccinated women, prevention of anal intraepithelial neoplasia of grade 2 or 3 is expected to reduce the risk of anal cancer among vaccinees. The qHPV vaccine also reduced the incidence of anal condyloma, a substantial added benefit of vaccination.
In summary, the qHPV vaccine is efficacious in reducing the incidences of persistent anal infection with HPV-6, 11, 16, or 18 and anal intraepithelial neoplasia associated with these HPV types. Unlike the screening and treatment of cervical intraepithelial neoplasia to reduce the risk of cervical cancer, there is currently no routine screening and treatment of anal intraepithelial neoplasia of grade 2 or 3 to reduce the risk of anal cancer. Vaccination may be the best long-term approach to reducing the risks of both anal cancer and anal condyloma.
Supported by grants from Merck and the National Institutes of Health (NIH/NCRR M01-RR-00079 and UL1 RR024131, to the University of California, San Francisco General Clinical Research Center, where some participants were studied).
Dr. Coutlee reports receiving funding through his institution from Merck Sharp & Dohme and providing expert testimony for GlaxoSmithKline and Qiagen. Dr. Ferris reports receiving funding through his institution, consulting fees and honoraria, and travel support from Merck Sharp & Dohme, serving as a board member for Merck Sharp & Dohme, and receiving consulting fees and honoraria from GlaxoSmithKline. Dr. Giuliano reports serving as a board member and consultant for and receiving grant support and honoraria from Merck Sharp & Dohme. Dr. Goldstone reports receiving grant support, consulting fees, and travel support from Merck Sharp & Dohme, serving as a consultant for and receiving grants and honoraria from Qiagen, receiving grants and travel support from the AIDS Malignancy Consortium, and receiving honoraria for the development of educational presentations and reimbursement for travel, accommodations, and meeting expenses from the American Social Health Association. Dr. Jessen reports receiving travel support from Merck Sharp & Dohme. Dr. Moreira reports receiving grant support through his institution, consulting fees, honoraria, travel support, and fees for review activities from Merck Sharp & Dohme and serving as a board member and consultant for Merck Sharp & Dohme. Dr. Palefsky reports receiving grant support through his institution, travel support, and fees for review activities through his institution from Merck Sharp & Dohme, serving as a board member and consultant and providing expert testimony for Merck Sharp & Dohme and receiving fees for these activities through his institution, serving as a member of the scientific advisory board of Pharmajet and receiving consulting fees for this activity through his institution, and serving as a member of the scientific advisory board of Aura Biosciences, for which he has received stock options. Dr. Stoler reports receiving consulting fees and honoraria through his institution from Merck Sharp & Dohme and serving as a consultant for Roche, Gen Probe, Qiagen, MTM Laboratories, Ventana, and Becton Dickinson. Drs. Garner, Guris, Haupt, Marshall, and Vuocolo and Mr. Radley report being employees of Merck Sharp & Dohme and owning stock or stock options. No other potential conflict of interest relevant to this article was reported.
A total of 602 men who have sex with men were enrolled (Table S1 in the Supplementary Appendix). Of these, 299 were vaccinated with qHPV vaccine and 299 with placebo. In all, 432 men who have sex with men (71.8%) had completed the 36-month follow-up period by the time the study was terminated. Roughly two thirds of participants were included in the per-protocol efficacy population, which was followed for a mean of 2.2 years after month 7.
The study groups were balanced with respect to age, race and ethnic group, region, smoking status, circumcision status, and sexual history (Table S2 in the Supplementary Appendix) as well as reason for discontinuation (data not shown). At baseline, 165 (27.4%) men were seropositive or HPV DNA–positive for HPV-6 or 11, 99 (16.4%) for HPV-16, and 68 (11.3%) for HPV-18. Among the 598 participants who received at least one dose of vaccine or placebo, 194 qHPV-vaccine recipients and 208 placebo recipients were eligible for the per-protocol efficacy analysis of HPV-6, 11, 16, or 18–related anal intraepithelial neoplasia end points (Fig. S1 in the Supplementary Appendix).
Efficacy in Preventing Anal Intraepithelial Neoplasia in Intention-to-Treat Population
In the intention-to-treat population, vaccine efficacy against anal intraepithelial neoplasia due to any HPV type was 25.7% (95% confidence interval [CI], -1.1 to 45.6) (Table 1Table 1Vaccine Efficacy against Anal Intraepithelial Neoplasia (AIN) and Anal Cancer in the Intention-to-Treat Population.). Efficacy against HPV-6, 11, 16, or 18–related anal intraepithelial neoplasia was 50.3% (95% CI, 25.7 to 67.2). Significant reductions in both anal intraepithelial neoplasia of grade 1 (49.6%; 95% CI, 21.2 to 68.4) and anal intraepithelial neoplasia of grade 2 or 3 (54.2%; 95% CI, 18.0 to 75.3) were seen in the intention-to-treat population. Efficacy ranged from 47.3 to 61.7% for anal intraepithelial neoplasia analyzed according to HPV type but was statistically significant only for HPV-6. Figure 1. shows the time to detection of HPV-6, 11, 16, or 18–related anal intraepithelial neoplasia in the per-protocol efficacy population and intention-to-treat population, as well as the time to detection of anal intraepithelial neoplasia related to any HPV type in the intention-to-treat population.
Efficacy in Preventing Anal Intraepithelial Neoplasia in Per-Protocol Population
Table 2. shows the primary efficacy data against anal intraepithelial neoplasia in the per-protocol efficacy population and efficacy according to HPV type and grade of anal intraepithelial neoplasia. HPV-6, 11, 16, or 18–related anal intraepithelial neoplasia developed in 5 participants in the vaccine group and 24 in the placebo group, with an observed efficacy of 77.5% (95% CI, 39.6 to 93.3). The vaccine was efficacious against both anal intraepithelial neoplasia of grade 1 (including condyloma) (73.0%; 95% CI, 16.3 to 93.4) and anal intraepithelial neoplasia of grade 2 or 3 (74.9%; 95% CI, 8.8 to 95.4). A total of 4 anal intraepithelial neoplasia lesions of grade 1 and 3 lesions of grade 2 or 3 related to HPV-6, 11, 16, or 18 developed in the vaccine group, as compared with 16 and 13, respectively, in the placebo group. Efficacy ranged from 65.5 to 100% for anal intraepithelial neoplasia analyzed according to HPV type but was statistically significant only for HPV-11. No cases of anal cancer developed in either study group.
Table S3 in the Supplementary Appendix shows the detection of HPV in swabs and biopsy samples from participants in whom anal intraepithelial neoplasia developed in association with more than one HPV type. In a post hoc case-assignment analysis performed in the per-protocol efficacy population to ascertain which of multiple HPV types detected was most likely to have caused the lesion, three lesions in the vaccine group, but none in the placebo group, were reassigned to nonvaccine types. The resulting recalculated efficacies against HPV-6, 11, 16, or 18–related anal intraepithelial neoplasia grade 1 and grade 2 or 3 were 91.1% (95% CI, 64.2 to 99.0) and 91.7% (95% CI, 44.6 to 99.8), respectively.
Efficacy in Preventing Persistent Anal HPV Infection
In the intention-to-treat population, use of the qHPV vaccine significantly reduced persistent infection with HPV-6, 11, 16, or 18, with an observed efficacy of 59.4% (95% CI, 43.0 to 71.4) (Table 3). The rate of infection with HPV-6, 11, 16, or 18 at any time was reduced by 48.5% (95% CI, 32.3 to 61.1). Significant reductions were found for persistent infection with each of the four HPV types, as well as detection of their DNA at any time.
Table 4 shows vaccine efficacy against persistent infection with HPV and detection of HPV DNA at any time in the per-protocol efficacy population. The reduction in persistent anal HPV-6, 11, 16, or 18 infection was 94.9% (95% CI, 80.4 to 99.4). Efficacy against all vaccine types was high, and for HPV-6, 16, and 18 it was significant. Efficacies against persistent HPV-16 and 18 infection were 93.8% (95% CI, 60.0 to 99.9) and 100% (95% CI, 51.5 to 100), respectively. Vaccinated participants had an 84.0% reduction (95% CI, 68.6 to 92.7) in detection of HPV-6, 11, 16, or 18 DNA at any time. Reductions in the rate of detection of DNA from each of the four HPV types were significant and ranged from 76.2 to 100%.
Vaccine efficacy against HPV-6, 11, 16, or 18–related anal intraepithelial neoplasia lesions among participants who were seropositive was 100% (95% CI, -26.2 to 100) in the subgroup without vaccine-type DNA on day 1 and 21.3% (95% CI, -94.2 to 69.1) in the subgroup with vaccine-type DNA on day 1 (Tables S5A and S5B in the Supplementary Appendix).
Table 5 presents the adverse events reported during the study period. The proportions of participants reporting adverse events were similar in the vaccine group and the placebo group. One or more adverse events were reported by 69.8% of qHPV-vaccine recipients and 70.6% of placebo recipients. The majority of events were local injection-site reactions, the rate of which was similar in the two groups. Few participants (1.3% in the vaccine group and 1.0% in the placebo group) reported having an injection-site adverse event that was “severe” (the worst possible classification). Approximately 18% of recipients in each group reported vaccine-related systemic adverse events. Details of systemic and injection-site adverse events are given in Table S4 in the Supplementary Appendix. No vaccine-related serious adverse events or deaths were reported in either group.
The trial was designed by the sponsor (Merck) in collaboration with three academic authors and an external data and safety monitoring board. The sponsor collated the data, monitored the conduct of the trial, performed statistical analyses, and coordinated the writing and revision of the manuscript among all the authors. All the authors were actively involved in the collection, analysis, and interpretation of the data; the decision to submit the manuscript for publication; and approval of the final version. The first draft was written by an academic author with contributions from another academic author and two industry authors. All the authors vouch for the completeness and accuracy of the analyses presented. The study was conducted in accordance with the protocol (Merck protocol 020), available with the full text of this article at NEJM.org.
The institutional review board at each participating center approved the protocol. All study participants gave written informed consent. Studies were conducted in conformity with applicable country or local requirements and informed-consent and other statutes or regulations regarding the protection of the rights and welfare of human subjects participating in biomedical research.
Between September 3, 2004, and August 29, 2008, we enrolled 3463 heterosexual men and 602 men who have sex with men in a randomized, placebo-controlled, double-blind study of the qHPV vaccine to prevent external genital lesions.12 Here, we describe a substudy of anal HPV infection and anal intraepithelial neoplasia in the men who have sex with men. We selected men who have sex with men for this substudy because the high incidence of anal infection and disease in this group was anticipated to allow for study completion within acceptable timelines.
The 602 men who have sex with men were enrolled in seven countries (Australia, Brazil, Canada, Croatia, Germany, Spain, and the United States), but the study was not powered to detect significant differences in vaccine efficacy or safety among countries. Inclusion criteria included an age of 16 to 26 years, five or fewer lifetime sexual partners, and engagement in insertive or receptive anal intercourse or oral sex with another boy or man within the past year. Exclusion criteria included a history or presence of clinically detectable anogenital warts or genital lesions suggesting other sexually transmitted diseases or an intra-anal lesion on anoscopy consistent with anal intraepithelial neoplasia or condyloma. Participants found to be HIV-positive before the first day of the study were excluded from the trial. Thirty-three participants diagnosed with HIV during the study were not withdrawn from the trial; they were referred for appropriate counseling and treatment and participated in all study procedures. Table S1 in the Supplementary Appendix (available at NEJM.org) provides further eligibility information.
Vaccine and Randomization
The qHPV L1 viruslike particle vaccine (Gardasil or Silgard, Merck) has been described previously.13 Men who have sex with men were randomly assigned, in a 1:1 ratio and according to a computer-generated schedule produced by the sponsor, to receive qHPV vaccine or placebo at day 1, month 2 (Ī3 weeks), and month 6 (Ī4 weeks). Vaccine or placebo was administered as a 0.5-ml injection in the deltoid muscle, generally on the same side of the body throughout the study. All investigators and site personnel, participants, monitors, and laboratory personnel remained unaware of the treatment assignments throughout the study, as did the sponsor's staff from the time of study onset through the time of the database lock for analysis.
Objectives and Study Measurements
Serum specimens for HPV serologic testing were obtained on study day 1 and month 7. Detailed anal examinations were scheduled for day 1 and months 7, 12, 18, 24, 30, and 36. At each visit, consecutive Dacron swabs were inserted into the anal canal to collect cells for anal cytologic analysis and HPV DNA,14 followed by a digital rectal examination and standard anoscopy. Participants underwent high-resolution anoscopy with biopsy of visible lesions if an abnormality was felt on digital rectal examination or seen on standard anoscopy, if anal cytologic testing showed atypical squamous cells of undetermined significance or more serious signs, or if HPV-related perianal lesions were histologically confirmed. All participants underwent high-resolution anoscopy and biopsy of any visible lesions at the exit visit.
Intra-anal samples were tested for HPV DNA with the use of multiplex polymerase-chain-reaction (PCR) assays, as described previously, to identify participants who were infected before enrollment or in whom new HPV infections developed during the study.10,11 Each biopsy-obtained thin section and each swab specimen was evaluated with the use of three primer-pair sets per HPV type, which amplified a portion of three separate open reading frames. Thin sections for which a specific HPV type was amplified in two or more PCR assays for the same HPV type were classified as HPV-positive for that type. All biopsy specimens were processed independently to prevent contamination of HPV DNA and were read in a blinded fashion, first by pathologists at the central laboratory for purposes of clinical management and then by a panel of pathologists for end-point adjudication. HPV testing of thin-section specimens was performed at the central laboratory.
To assess vaccine safety, participants used vaccination report cards to record oral temperature and adverse events occurring at the injection site 1 to 5 days after each vaccination and systemic and serious adverse events occurring 1 to 15 days after each vaccination. In addition, all serious adverse events, including those considered related to the vaccine or a study procedure by the investigators, were recorded, as were all deaths.
Study End Points
Efficacy was measured in the per-protocol efficacy population: participants who were seronegative and had HPV DNA–negative swab and biopsy specimens at day 1 for relevant vaccine types, were negative for vaccine-type DNA through month 7, and did not violate the protocol. Case counting in this population commenced at month 7.
The intention-to-treat population consisted of participants who were or were not seropositive or DNA-positive for the vaccine HPV types at enrollment, received at least one dose of vaccine or placebo, and returned for follow-up. Case counting commenced after day 1.
The prespecified primary efficacy end point was HPV-6, 11, 16, or 18–related anal intraepithelial neoplasia or anal cancer. End points were determined to have occurred in a biopsy specimen if the following were true: a consensus diagnosis by the pathology panel of anal intraepithelial neoplasia grade 1 (including condyloma), grade 2, or grade 3 or anal cancer; and detection of HPV-6, 11, 16, or 18 DNA by means of PCR assay in a section adjacent to the section used for histologic diagnosis. Some participants may have contributed more than one lesion to the analysis.
The prespecified secondary efficacy end point of persistent HPV infection was defined as detection of the same HPV type (HPV-6, 11, 16, or 18) in an anal swab or biopsy specimen collected during two or more consecutive visits 6 months or more (Ī1 month) apart. Participants for whom HPV-6, 11, 16, or 18 DNA was detected in any swab or biopsy specimen during at least one visit were included in the end point of DNA detection at any time during the study.
The primary efficacy end point was the incidence of HPV-6, 11, 16, or 18–related anal intraepithelial neoplasia or anal cancer. Assuming a true vaccine efficacy for this end point of 85%, a lower bound for the confidence interval for vaccine efficacy greater than 0%, a one-sided α value of 0.025, and equal duration of follow-up in the vaccine and placebo groups, we calculated that 17 cases of the primary efficacy end point among men who have sex with men would be required in the per-protocol efficacy population for at least 90% power to declare the vaccine efficacious. For the purpose of calculating the necessary sample size, we conservatively assumed that the study would continue until the 17 cases had been observed in the placebo group (i.e., vaccine efficacy is 100%).
Multiple HPV types were sometimes detected in biopsy specimens of lesions. A post hoc case-assignment analysis was performed in the per-protocol efficacy population to identify the HPV type most likely to have caused a given lesion. The methods used to assign a specific HPV type to a lesion is described in the Supplementary Appendix