Costa Rica; HPV; HPV vaccine; vulvar human papillomavirus vaccine.

Transcription

1 MAJOR ARTICLE Reduced Prevalence of Vulvar HPV16/18 Infection Among Women Who Received the HPV16/18 Bivalent Vaccine: A Nested Analysis Within the Costa Rica Vaccine Trial Krystle A. Lang Kuhs, 1 Paula Gonzalez, 2,3 Ana Cecilia Rodriguez, 2 Leen-Jan van Doorn, 4 Mark Schiffman, 1 Linda Struijk, 4 Sabrina Chen, 5 Wim Quint, 4 Douglas R. Lowy, 1 Carolina Porras, 2 Corey DelVecchio, 5 Silvia Jimenez, 2 Mahboobeh Safaeian, 1 John T. Schiller, 1 Sholom Wacholder, 1 Rolando Herrero, 3 Allan Hildesheim, 1 and Aimée R. Kreimer 1 ; for the Costa Rica Vaccine Trial Group 1 National Cancer Institute, NIH, Bethesda, Maryland; 2 Proyecto Epidemiológico Guanacaste, Fundación INCIENSA, Costa Rica; 3 Prevention and Implementation Group, International Agency for Research on Cancer, Lyon, France; 4 DDL Diagnostic Laboratory, Rijswijk, The Netherlands; and 5 Information Management Services, Calverton, Maryland Background. Vaccine efficacy (VE) against vulvar human papillomavirus (HPV) infection has not been reported and data regarding its epidemiology are sparse. Methods. Women (n = 5404) age present at the 4-year study visit of the Costa Rica Vaccine Trial provided vulvar and cervical samples. A subset (n = 1044) was tested for HPV DNA (SPF 10 /LiPA 25 version 1). VE against 1-time detection of vulvar HPV16/18 among HPV vaccinated versus unvaccinated women was calculated and compared to the cervix. Prevalence of and risk factors for HPV were evaluated in the control arm (n = 536). Results. Vulvar HPV16/18 VE (54.1%; 95% confidence interval [CI], 4.9% 79.1%) was comparable to cervix (45.8%; 95% CI, 6.4% 69.4%). Vulvar and cervical HPV16 prevalence within the control arm was 3.0% and 4.7%, respectively. Independent risk factors for vulvar HPV were similar to cervix and included: age (adjusted odds ratio [aor] 0.5 [95% CI,.3.9] 28 vs 22 23]); marital status (aor 2.3 [95% CI, ] single vs married/living-asmarried); and number of sexual partners (aor 3.6 [95% CI, ] 6 vs 1). Conclusions. In this intention-to-treat analysis, VE against vulvar and cervical HPV16/18 were comparable 4 years following vaccination. Risk factors for HPV were similar by anatomic site. Clinical Trials Registration. NCT Keywords. Costa Rica; HPV; HPV vaccine; vulvar human papillomavirus vaccine. Vulvar cancer is rare, accounting for approximately 4% of gynecological malignancies worldwide [1]. However, the incidence of vulvar cancer has been increasing, particularly in young women [2, 3]. In the United States Received 14 March 2014; accepted 18 June 2014; electronically published 23 June Presented in part: 29th International Papillomavirus Conference, Seattle, Washington, August Correspondence: Krystle A. Lang Kuhs, PhD, MPH, Infections and Immunoepidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, 9609 Medical Center Dr, Rm 6-E210, Bethesda, MD (krystle.kuhs@ nih.gov). The Journal of Infectious Diseases 2014;210: Published by Oxford University Press on behalf of the Infectious Diseases Society of America This work is written by (a) US Government employee(s) and is in the public domain in the US. DOI: /infdis/jiu357 from 2000 to 2009, the incidence of vulvar cancer has significantly increased among both white and African American women; 1.4% and 0.9% annual percentage change, respectively [3]. This increase is believed to be due to a rise in human papillomavirus (HPV) infection [2, 4 6]. Worldwide, it is estimated that HPV infection is the causative agent in approximately 43% of vulvar cancers [7]. HPV 16 and 18 infections combined account for 93% of HPV-associated vulvar cancers [8]; thus, prophylactic vaccination has the potential to reduce the incidence of vulvar cancer. Two vaccines prevent HPV 16 and 18 infections at the cervix: the bivalent HPV16/18 vaccine Cervarix, GlaxoSmithKline Vaccines [GSK]) [9] and the quadrivalent HPV6/11/16/18 vaccine (Gardasil, Merck and 1890 JID 2014:210 (15 December) Lang Kuhs et al

2 Co, Inc) [10]. Data presented in abstract form for the bivalent vaccine reported a 75% reduction in HPV16/18-associated vulvar/vaginal precursor lesions (a combined endpoint) among vaccinated women compared to controls [11]. For the quadrivalent vaccine, high (>95%) vaccine efficacy (VE) against HPV16/ 18-associated vulvar intraepithelial neoplasia grades 1, 2, and 3 was demonstrated [10], garnering an indication from the US Food and Drug Administration (FDA) for prevention of vulvar cancer. VE against HPV16/18 infections at the vulva has yet to be reported. Efficacy against HPV16/18-associated vulvar lesions was demonstrated by both pharmaceutical vaccine trials using HPV DNA testing of vulvar lesions, which represented a small subset of the total trial population [10, 11]. Given that VE against HPV-associated lesions is the result of prevention of HPV infection, we reasoned that use of a virologic endpoint for the estimation of vulvar VE would provide additional novel informationonanintermediateendpointonthepathwayto cancer. Likewise, perhaps due to the rarity of HPV-associated vulvar cancer, only 6 studies have been published on the epidemiology of vulvar HPV infection in cancer-free women [12 17] and only 1 study reported risk factors associated with high-risk HPV infection of the vulva [17]. To address the limited knowledge of vulvar VE as well as the epidemiology of vulvar HPV infection, we implemented a vulvar specimen collection at the final (year 4), randomized blinded study visit of the Costa Rica Vaccine Trial (CVT). The goals of this value-added analysis were to (1) estimate the reduction of prevalent HPV16/18 infections at the vulva compared to cervix among HPV vaccinated women compared to controls 4 years after vaccination, and (2) characterize the prevalence of and risk factors for vulvar HPV infection in the control arm of this study population. METHODS Subject Participants and Study Design Women included in the present evaluation are participants in both arms of a double-blind, controlled, randomized, phase III study of the efficacy of an HPV16/18 virus-like particle vaccine conducted in CVT. This trial was designed to evaluate VE against persistent cervical HPV16/18 infection and precancerous lesions, as previously described [18]. Inclusion criteria were: age years (inclusive) at entry, generally healthy as determined by medical history and a physical examination, not pregnant or breastfeeding, using contraception (including abstinence) during the vaccination period, and willing to provide written informed consent. The trial was reviewed and approved by human subjects review committees of INCIENSA (Instituto Costarricense de Investigacion y Enseñanza en Nutrición y Salud) in Costa Rica and the National Cancer Institute in the United States. All subjects provided informed consent. Procedures At enrollment, women were blindly randomized to receive the HPV bivalent vaccine (Cervarix, GSK Vaccines) or a control hepatitis A vaccine (modified Havrix, GSK Vaccines) [19, 20]. At enrollment and at each annual visit, a questionnaire regarding the woman s education, marital status, sexual and reproductive history, and smoking habits was administered; a pelvic examination was performed on sexually experienced women, with collection of cervical cells for liquid-based cytology and HPV testing; and blood was collected for HPV16/18 serology. Women with low-grade cervical cytologic abnormalities were evaluated semiannually and those with high-grade or persistent low-grade cervical abnormalities were referred for colposcopy and treatment as needed [18]. A medical history was conducted during enrollment and during each follow-up visit as part of the process of assessment of adverse events, and information regarding medical conditions (including those that affect the female reproductive tract) was collected by study clinicians at each follow-up visit. At the 4-year study visit, an additional questionnaire regarding sexual history (including oral and anal sex) was administered, and all sexually experienced women were asked to provide a vulvar, cervical, and anal specimen, which were collected by study clinicians. Specimen Collection Prior to insertion of the speculum for cervical specimen collection, a vulvar sample was collected by swabbing the mucosal surface of each labia minora 4 times (for a total of 8 swipes using a single Dacron swab). The vulvar swab was placed in a vial of 1 ml PreservCyt and frozen in liquid nitrogen. Cervical specimens were collected with a Cervex-Brush (Rovers Medical Devices) by firmly rotating the brush 5 times 360 around the cervical os. The Cervex-Brush was placed in 20 ml PreservCyt, and 0.5 ml aliquots for HPV polymerase chain reaction (PCR) testing were prepared and frozen at the local laboratory following PCR-safe procedures. HPV DNA Testing A total of 5404 women agreed to both vulvar and cervical specimen collection at the 4-year study visit (when women were between the ages of 22 29). All 5404 cervical samples were tested for HPV DNA. The first 1044 of the 5404 vulvar samples collected were tested for HPV DNA; additional samples were not tested due to financial constraints. The same HPV DNA testing procedures were used for both the cervical and vulvar specimens, and all samples were tested within the same laboratory. DNA extraction via the MagNAPure LC DNA isolation procedure (Roche Diagnostics) was used. Extracted DNA samples were tested for HPV DNA using the HPV SPF 10 PCR-DEIA (DNA enzyme immunoassay)-lipa 25 (line probe assay) version 1 system (Labo Biomedical Products, Rijswijk, The Netherlands) [21, 22]. Briefly, this broad-spectrum PCR-based HPV Reduced Vulvar HPV16/18 With Vaccination JID 2014:210 (15 December) 1891

3 Figure 1. Trial profile. Abbreviation: HPV, human papillomavirus. a Four women received discordant vaccines (1 woman was enrolled twice and received 3 doses of each vaccine and 3 women received 2 doses of 1 vaccine and 1 dose of the other vaccine). For the aim of this analysis, the women were assigned to the group for which the first dose was given. b Data obtained at the 4-year study visit at which vulvar specimens were obtained were used to create the full analytic cohort. DNA testing system uses SPF 10 primers to amplify and a DEIA to detect at least 69 HPV genotypes and the LiPA 25 line probe assay to genotype carcinogenic and noncarcinogenic HPVs. To increase the sensitivity of type-specific detection of HPV16 and 18 using the SPF 10 DEIA system, all specimens positive for HPV DNA using SPF 10 DEIA but negative for HPV16 or HPV18 by LiPA 25 were additionally tested with type-specific primers [23]. Statistical Analysis General characteristics from the 4-year postvaccination visits were compared between women who had (1) a vulvar sample collected and tested, (2) a vulvar sample collected and not tested, and (3) refused vulvar sample collection. Groups were compared using a χ 2 test. Restricting to women who had a vulvar sample collected and tested (group 1 above), both enrollment and 4-year postvaccination characteristics between women in the vaccine and control arms were compared using a χ 2 test (Supplementary Table 1). Our approach to calculating VE at extra-cervical sites at the 4- year study visit has been previously published [19, 24]. We evaluated prevalence of vulvar HPV16 or 18 infections approximately 4 years after vaccination (defined as detection of either HPV16 or HPV18 or both at the 4-year study visit); cervical HPV16 or 18 infections among the same women at the same time point were evaluated for comparison purposes. Our intention-to-treat analytic cohort (Figure 1) included all women who had vulvar specimens collected and had HPV results available; no exclusions were made based on enrollment HPV DNA positivity, HPV serostatus, or number of vaccine doses received. While we desired to create an additional HPV-naive analytic cohort (defined as HPV negative at the cervix at baseline), we did not have adequate power to pursue this analysis. For each arm, the prevalence of HPV16/18 measured 4 years postvaccination were expressed as the number of infected women per 100 women vaccinated (stratified by HPV vs control vaccine); asymptotic confidence intervals (95% confidence interval [CI]) are reported. The complement of the ratios of the prevalence for the HPV and control arms comprised the VE estimates. Exact confidence intervals for VE were calculated based on the binomial distribution of the number of events in the HPV arm among the 1892 JID 2014:210 (15 December) Lang Kuhs et al

4 total number of events in the HPV and control arms [25, 26]. Vulvar and cervical VE estimates were calculated and compared by including an arm*anatomical site interaction variable in a generalized estimating equation model [27] and evaluating whether the β coefficient for the interaction variable varied significantly from zero, a nonsignificant P value would be interpreted as the 2 VE estimates are not significantly different. The epidemiology of vulvar versus cervical HPV infection was evaluated in the control arm. Prevalence of vulvar and cervical HPV were described separately among all women. Analyses for grouped HPV infections (any HPV, carcinogenic, noncarcinogenic HPV) were conducted at the woman-level; for analyses by individual HPV type, a woman could only contribute once to each analysis. A κ statistic was calculated for vulvar and cervical HPV infection for each of the 25 HPV types. McNemar s test was used to determine κ significance. In cases of less than 5 events per cell, χ 2 tests were used instead. Risk factors for vulvar HPV infection were evaluated in the control arm only using logistic regression analysis. Variables evaluated in univariate modeling included: age, education, marital status, smoking status, age at first vaginal intercourse, lifetime number of sexual partners, and total number of years sexually active. The association between vulvar HPV and medical conditions of the female reproductive tract was also evaluated. To evaluate the independent determinants of vulvar HPV, risk factors that were significantly associated with vulvar HPV in the univariate analysis or considered important based on the literature (ie, age and lifetime number of sexual partners) were included in multivariate models as adjusted odds ratios (aors). The same analyses were conducted using cervical HPV as the outcome. For all analyses, a P value <.05 (2-tailed) was considered statistically significant. RESULTS Participant Characteristics Of the 7466 women randomized (Figure 1), 6351 attended the 4-year study visit (3180 HPV arm; 3171 Control arm). Sexually naive women were not eligible for vulvar specimen collection (n = 382). Of the eligible women (n = 5969), 565 (297 HPV arm; 268 Control arm) women refused vulvar specimen collection and 5404 accepted (90.5%). Of the 5404 women that provided a vulvar specimen, the first 1044 samples were tested for HPV DNA (508 HPV arm; 536 Control arm). Compared to women whom had a vulvar sample collected (n = 5404), women who refused vulvar sample collection (n = 565) appeared generally less sexually experienced, a finding we expected based on previous work (Table 1) [19]. Participant characteristics were similar for women for whom the vulvar sample was HPV tested (n = 1044) and women for whom the vulvar sample was not HPV tested (n = 4360) with regard to age, education level, marital status age at first vaginal intercourse, total number of years sexually active and cervical HPVstatus.However,womeninouranalyticcohort(n= 1044) were more likely to be never smokers (82.9% vs 78.8%) and to have fewer number of lifetime sexual partners (median number of sexual partners 2 vs 3) compared to women for whom the vulvar sample was not HPV tested. Distributions of pre- and postrandomization characteristics were similar in the HPV and control arms; all P values were >.3 (Supplementary Table 1). Reduction of Prevalent HPV16/18 Infection Among HPV-Vaccinated Women at the 4-Year Study Visit Significant VE against prevalent HPV16/18 infection at the 4-year study visit was observed both at the vulva and cervix. Vulvar VE was 54.1% (95% CI, ; 10 and 23 events in the HPV and control arms, respectively); the corresponding cervical VE was 45.8% (95% CI, ; 19 and 37 events in the HPV and control arms, respectively). Vaccine efficacy at the vulva did not differ significantly from vaccine efficacy at the cervix; p for anatomic-site interaction = 0.6 (Table 2). In order to test the representativeness of our subset of 1044 women compared to the full cohort of women who provided a vulvar sample, we compared cervical VE for both cohorts. While participant characteristics between the 2 groups were similar (Table 1), cervical VE for the full cohort (n = 5404) was higher than the subset (n = 1044); 71.7% (95% CI, ) versus 45.8% (95% CI, ), respectively, albeit with overlapping confidence intervals. Vulvar HPV Prevalence and Type Distribution To further our understanding of vulvar HPV epidemiology, we conducted additional prevalence and risk factor analyses in the control arm only. Overall prevalence of any HPV type at the vulva was 29.5% (95% CI, ; 158 HPV infections detected) (Table 3). The prevalence of carcinogenic and noncarcinogenic HPV types was 16.8% (95% CI, ) and 17.2% (95% CI, ), respectively, the total of which exceeds the overall prevalence (29.5%) due to detection of multiple infections. The most common carcinogenic HPV types were HPV52 (3.4% [95% CI, ]), HPV56 (3.4% [95% CI, ]) and HPV16 (3.0% [95% CI, ]). Single vulvar HPV infections were detected in 20.5% (95% CI, ) of participants, while 2 or more types were detected in 9.0% (95% CI, ). Overall prevalence of any HPV type at the cervix was similar to that of the vulva (34.1% [95% CI, ; 183 HPV infections detected]) (Table 3). The prevalence of carcinogenic and noncarcinogenic HPV types at the cervix was 22.2% (95% CI, ) and 17.7% (95% CI, ), respectively. The most common carcinogenic type was HPV16 (4.7% [95% CI, ]). A κ statistic was computed in order to determine the correlation between vulvar and cervical HPV16 infection within the same women in the control arm only (n = 536). Reduced Vulvar HPV16/18 With Vaccination JID 2014:210 (15 December) 1893

5 Table 1. Participant Characteristics at the 4-Year Study Visit Accepted Vulvar Sample Collection (n = 5404) Accepted Vulvar Sample Collection (N = 5404) Refused Vulvar Sample Collection (N = 565) Sample HPV Tested (N = 1044) Sample Not HPV Tested (N = 4360) Characteristics N (%) N (%) P Value a N (%) N (%) P Value a Age (y).8.3 Median (IQR) 26 (24 28) 26 (24 28) 26 (24 28) 26 (24 28) (21.4) 131 (23.2) 218 (20.9) 941 (21.6) (26.6) 150 (26.5) 259 (24.8) 1177 (27.0) (23.9) 129 (22.8) 269 (25.8) 1023 (23.5) (28.1) 155 (27.5) 298 (28.5) 1219 (27.9) Education level y 1484 (27.5) 152 (26.9) 304 (29.1) 1180 (27.1) 7 9 y 1027 (19.0) 99 (17.5) 179 (17.1) 848 (19.4) 10 y* 1476 (27.3) 146 (25.8) 289 (27.7) 1187 (27.2) University 1417 (26.2) 168 (29.8) 272 (26.1) 1145 (26.3) Marital status Married or living as 3552 (65.7) 369 (65.3) 713 (68.3) 2839 (65.1) married* Single 1501 (27.8) 160 (28.3) 276 (26.4) 1225 (28.1) Divorced/Separated/ 351 (6.5) 36 (6.4) 55 (5.3) 296 (6.8) Widowed Smoking status Never* 4303 (79.6) 485 (85.8) 866 (82.9) 3437 (78.8) Former** 722 (13.4) 54 (9.6) 124 (11.9) 598 (13.7) Current 379 (7.0) 26 (4.6) 54 (5.2) 325 (7.5) Age at first vaginal < intercourse (years) Median (IQR) 17 (15 19) 18 (16 20) 17 (15 19) 17 (15 19) (27.1) 210 (37.2) 312 (29.9) 1153 (26.4) (30.6) 159 (28.1) 305 (29.2) 1350 (31.0) 16* 2284 (42.3) 196 (34.7) 427 (40.9) 1857 (42.6) Lifetime number of < sexual partners Median (IQR) 3 (1 5) 2 (1 4) 2 (1 4) 3 (1 5) (26.7) 203 (36.0) 326 (31.2) 1117 (25.6) (56.9) 303 (53.5) 578 (55.4) 2497 (57.3) (16.4) 59 (10.5) 140 (13.4) 746 (17.1) Total number of years.04.8 sexually active Median (IQR) 9 (6 11) 8 (5 11) 9 (6 11) 9 (6 11) 0 7 y* 2089 (38.7) 248 (43.9) 410 (39.2) 1679 (38.5) 8 10 y 1687 (31.2) 168 (29.7) 318 (30.5) 1369 (31.4) 11 y 1628 (30.1) 149 (26.4) 316 (30.3) 1312 (30.1) Cervical HPV status b.05.3 Negative* 3493 (64.6) 389 (68.8) 689 (66.0) 2804 (64.3) Positive 1911 (35.4) 176 (31.2) 355 (34.0) 1556 (35.7) Participant characteristic were compared for women who accepted vulvar sample collection (N = 5404) versus women who refused vulvar sample collection (N = 565). Among women who agreed to vulvar sample collection (N = 5404), participant characteristics were compared for women for whom the vulvar sample was HPV tested (N = 1044) versus not HPV tested (N = 4360). Missing values accounted for less than 2 percent of the full cohort; yet, when a value was missing: *includes women with a missing value, those who refused to respond, and those who responded don t know for this variable; **includes women who did not have missing values for a related variable above, yet had a missing value, refused to respond or responded don t know for this variable. Abbreviations: HPV, human papillomavirus; IQR, interquartile range. a χ 2 P value. b Any HPV infection detected (carcinogenic and/or noncarcinogenic) JID 2014:210 (15 December) Lang Kuhs et al

6 Table 2. Estimated Vaccine Efficacy Against Vulvar and Cervical HPV16 and/or 18 Infections in the Intention-to-Treat Analytic Cohort at the 4-Year Study Visit Anatomic Site Arm No. of Women No. of HPV16/18 Infections HPV16/18 Prevalence (95% CI) HPV Vaccine Efficacy (95% CI) a Vulvar HPV ( ) 54.1% (4.9% 79.1%) Control ( ) Cervix HPV ( ) 45.8% (6.4% 69.4%) Control ( ) Analysis included all women who had vulvar specimens collected and had HPV results available; no exclusions were made based on number of vaccine doses received, HPV DNA positivity or HPV serostatus at baseline. Abbreviations: CI, confidence interval; HPV, human papillomavirus; VE, vaccine efficacy. a Vulvar and cervical VE estimates were calculated and compared by including an arm*anatomical site interaction variable in a generalized estimating equation (GEE) model; p for interaction by anatomical site = 0.6. Table 3. Prevalence of Vulvar and Cervical HPV Types at the 4-Year Study Visit Among the 536 Young Adult Women From Costa Rica in the Control Arm Only Who Provided Both a Vulvar and Cervical Specimen for HPV DNA Testing, Stratified by Anatomic Site Vulva Cervix No. of Prevalence, % of No. of Prevalence, % of Test Result HPVs Detected All Women (95% CI) HPVs Detected All Women (95% CI) Any a ( ) ( ) Carcinogenic b ( ) ( ) ( ) ( ) (.6 2.9) ( ) ( ) ( ) (.0 1.0) (.3 2.2) (.9 3.4) (.9 3.4) ( ) (.9 3.4) (.3 2.2) (.5 2.7) (.9 3.4) ( ) ( ) ( ) ( ) ( ) (.3 2.2) (.6 2.9) (.0 1.3) (.4 2.4) Noncarcinogenic c ( ) ( ) (.3 2.2) (.2 1.9) (.0 1.3) (.0 1.3) (.0 1.3) (.3 2.2) (.1 1.6) (.4 2.4) (.4 2.4) (.9 3.4) (.4 2.4) ( ) ( ) ( ) ( ) ( ) ( ) 68/ ( ) ( ) ( ) ( ) ( ) ( ) Abbreviations: CI, confidence interval; HPV, human papillomavirus. a HPV 6, 11, 16, 18, 31, 33, 34, 35, 39, 40, 42, 43, 44, 45, 51, 52, 53, 54, 56, 58, 59, 66, 68/73, 70, and 74. b HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59. c HPV 6, 11, 34, 40, 42, 43, 44, 53, 54, 66, 68/73, 70, and 74. Reduced Vulvar HPV16/18 With Vaccination JID 2014:210 (15 December) 1895

7 Table 4. Univariate and Multivariate Analyses of Risk Factors for Detection of Any Vulvar HPV at the 4-Year Study Visit Among Young Adult Women From Costa Rica (Control Arm Only) Control Arm (N = 536) HPV Positivity, Characteristic No. of Women No. (%) of Women OR (95% CI) aor (95% CI) Age (y) (41.2) Ref Ref (31.3) 0.7 (.4 1.1) 0.7 (.4 1.3) (22.1) 0.4 (.2.7) 0.4 (.2.7) (26.9) 0.5 (.3.9) 0.5 (.3.9) p for trend Education level 6 y (25.9) Ref 7 9 y (29.2) 1.2 (.7 2.1) 10 y* (27.8) 1.1 (.7 1.8) University (35.2) 1.6 (.9 2.6) p for trend 0.1 Marital status Married or living as married* (23.1) Ref Ref Single (45.5) 2.8 ( ) 2.3 ( ) Divorced/Separated/Widowed (40.0) 2.2 ( ) 1.9 (.8 4.5) Smoking status a Never* (26.7) Ref Former** (45.5) 2.3 ( ) Current (44.4) 2.2 ( ) p for trend Age at first vaginal intercourse (y) (31.4) Ref (26.5) 0.8 (.5 1.3) 16* (30.2) 0.9 (.6 1.5) p for trend 0.9 Lifetime number of sexual partners (15.4) Ref Ref (34.8) 2.9 ( ) 2.5 ( ) (42.3) 4.0 ( ) 3.6 ( ) p for trend < < Total number of years sexually active 0 7 y* (33.3) Ref 8 10 y (28.2) 0.8 (.5 1.2) 11 y (26.1) 0.7 (.4 1.1) p for trend 0.1 Missing values accounted for less than 2 percent of the full cohort; yet, when a value was missing: *includes women with a missing value, those who refused to respond, and those who responded don t know for this variable; **includes women who did not have missing values for a related variable above, yet had a missing value, refused to respond or responded don t know for this variable. Statistically significant variables in the multivariate model are bolded. ORs were adjusted for all the variables in bold. Abbreviations: aor, adjusted odds ratio; CI, confidence interval; HPV, human papillomavirus; OR, odds ratio; Ref, reference. a Smoking was not significant in the multivariate model and was therefore dropped from the final model. A total of 16 vulvar and 25 cervical HPV16 infections were detected. Of the 16 women with a vulvar HPV16 infection, 13 women had a concurrent HPV16 cervical infection. Twelve of 25 women had cervical HPV16 infection only. Moderate correlation was observed between vulvar and cervical HPV16 infection; κ =0.6, P =.02. For each of the additional 24 HPV types tested within the control arm, κ ranged from 0.2 to JID 2014:210 (15 December) Lang Kuhs et al

8 Risk Factor Analysis The evaluation of risk factors for any vulvar or cervical HPV was again restricted to the control arm (Table 4). In the multivariate analysis, age was inversely associated with vulvar HPV infection (the oldest women [age 28] were half as likely to be vulvar HPV positive compared to the youngest women [age 22 23]); additional independent risk factors for any vulvar HPV infection included marital status (aor, 2.3 [95% CI, ] single compared to married/living as married), number of sexual partners (aor, 3.6 [95% CI, ] 6 partners compared to 1). Following adjustment, smoking lost statistical significance (aor, 1.3 [95% CI,.6 3.1] for current vs nonsmokers) and was therefore dropped from the final multivariate model. When any cervical HPV infection was used as the outcome in place of any vulvar HPV infection, risk factors as well as the magnitude of risk were similar to that of the vulva (data not shown). Medical conditions affecting the vulva that were explored but were not found to be associated with HPV infection included: trichomoniasis; other inflammation of the vagina and vulva; absent, scanty, and rare menstruation; excessive, frequent, and irregular menstruation; other abnormal uterine and vaginal bleeding; and pain and other conditions associated with female genital organs and menstrual cycle. DISCUSSION This analysis of data from our randomized, community-based HPV vaccine trial in Costa Rica shows that the bivalent HPV vaccine is efficacious against vulvar HPV16/18 infections in young women measured 4 years after vaccination, and that the observed protection at the vulva is comparable to that observed at the cervix. Due to this being an intention-to-treat analysis conducted at 1 time point within a population that includes HPV-exposed women, vulvar VE estimates in our study are lower than what would be expected among the target vaccination population of HPV-naive adolescents where cervical VE is estimated to approach 100% [28]. However, given that our estimates of cervical and vulvar VE were comparable, we expect based on our study that vulvar VE among HPV-naive adolescents would be as high as cervical VE. Based on the cumulative work to date in our trial, we now understand the HPV vaccine protects against HPV infections at the anatomic sites where HPV causes cancer in women, including the cervix [20], vulva, anus [19], and oral region [24]. In this study, we utilized HPV infection as the outcome due to the rarity of vulvar dysplasia. Thus, understanding the epidemiology of vulvar HPV infection provides an important framework for interpreting our vaccine efficacy results. There have been 6 studies published on the epidemiology of vulvar HPV infection in cancer-free women [12 17]. In our study of young women within the control arm, detection of vulvar HPV was relatively common and mirrored that of the cervix; overall prevalence of HPV among women approached 30%. Both carcinogenic and noncarcinogenic HPV were similarly prevalent (approximately 17%); HPV16 was one of the more common of the carcinogenic types detected (3.0%). The strongest risk factor for vulvar HPV in our study and in a comparable study by Howell-Jones et al [17] was higher number of sexual partners, a finding in line with what is known for cervical HPV epidemiology. Age was another important risk factor for vulvar HPV infection: our point prevalence of vulvar HPV16 was highest among our youngest women (5.9% among year olds compared to 1.9% among women 28 years old), which is comparable to previously reported prevalence estimates in the literature [12 17]. Thus, it appears that age-specific vulvar HPV prevalence parallels that of cervical HPV prevalence, which peaks shortly after sexual debut. There are important considerations in our study. First, in contrast to using precancerous lesions as an outcome, or even persistent HPV infection, 1-time HPV detection is unable to distinguish true vulvar HPV infection from viral deposition from either recent sexual contact or, cervical HPV infection. In this study, half of the 25 women with cervical HPV16 infection had concurrent vulvar infection detected, suggesting that viral deposition from a cervical infection may not be as common as expected. Measures taken to reduce contact of the vulvar specimen with cervical fluids (the vulvar sample was collected prior to insertion of the speculum for cervical specimen collection) may have reduced this potential problem. Second, while we wanted to conduct an additional analysis of vulvar VE in an HPV-naive cohort (based on cervical HPV status at initial vaccination), there were not adequate endpoints to pursue this analysis. Third, not all vulvar samples were HPV tested. The descriptive characteristics of the tested subset were mostly similar to that of the full cohort with vulvar specimen collection; yet, notable differences in cervical VE between the 2 groups suggest the presence of residual variability not captured by our questionnaire. If the tested subset was less risky compared to the full cohort (evidenced by having fewer lifetime sexual partners), then the vulvar VE estimates could have been inflated by the majority of infections detected being incident and therefore preventable by vaccination. If our analytic subset was more risky (evidenced by the higher cervical VE observed in the full cohort), then the opposite could be observed: more prevalent infections at baseline would result in a decrement in the observed VE [29]. While our subset of 1044 may have not been representative of the full cohort, balance was observed between arms; thus, we believe this work yields valid estimates. In the subset, VE estimates for the cervix and the vulva were similar, which we interpret to mean that the vaccine comparably protects against HPV16/18 infections at both anatomic sites. Our findings suggest that the bivalent HPV vaccine protects against vulvar HPV16 and 18 infections. Findings presented here and in the literature suggest that the incidence of HPV Reduced Vulvar HPV16/18 With Vaccination JID 2014:210 (15 December) 1897

9 infection and resultant cancers will decrease at several anatomic sites in women who receive the prophylactic HPV vaccines before exposure to the virus [20]. Unfortunately, among unvaccinated women, vulvar HPV infection will remain common. Supplementary Data Supplementary materials are available at The Journal of Infectious Diseases online ( Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author. Notes Acknowledgments. We would like to extend a special thanks to the women of Guanacaste and Puntarenas, Costa Rica, who gave of themselves in participating in this effort. We also acknowledge the tremendous effort and dedication of the staff in Costa Rica involved in this project. In the United States, we extend our appreciation to the team from Information Management Services (IMS) responsible for the development and maintenance of the data system used in the trial and who serve as the data management center for this effort. We would like to specifically acknowledge the invaluable contributions made by Jean Cyr, Julie Buckland, John Schussler, and Brian Befano. We thank the members of the Data and Safety Monitoring Board charged with protecting the safety and interest of participants in our trial (Steve Self, Chair, Adriana Benavides, Luis Diego Calzada, Ruth Karron, Ritu Nayar, and Nancy Roach) and members of the external Scientific HPV Working Group who have contributed to the success of our efforts over the years (Joanna Cain, Chair, Diane Davey, David DeMets, Francisco Fuster, Ann Gershon, Elizabeth Holly, Silvia Lara, Henriette Raventós, Wasima Rida, Luis Rosero-Bixby, Kristen Suthers, and Sarah Thomas). Protocol available at Cervarix and Havrix are trademarks of the GlaxoSmithKline group of companies. Gardasil is a registered trade mark of Merck & Co, Inc. Registered with Clinicaltrials.gov NCT Investigators in the Costa Rica Vaccine Trial (CVT) group. Proyecto Epidemiológico Guanacaste, Fundación INCIENSA, San José, Costa Rica Mario Alfaro (cytopathologist), Manuel Barrantes (field supervisor), M. Concepción Bratti (co-investigator), Fernando Cárdenas (general field supervisor), Bernal Cortés (specimen and repository manager), Albert Espinoza (head, coding and data entry), Yenory Estrada ( pharmacist), Paula González (co-investigator), Diego Guillén ( pathologist), Roland Herrero (coprincipal investigator), Silvia E. Jiménez (trial coordinator), Jorge Morales (colposcopist), Luis Villegas (colposcopist), Lidia Ana Morera (head studynurse),elmerpérez(fieldsupervisor), CarolinaPorras(co-investigator), Ana Cecilia Rodríguez (co-investigator), Libia Rivas (clinical coordinator). University of Costa Rica, San José, Costa Rica Enrique Freer (director, HPV diagnostics laboratory), José Bonilla (head, HPV immunology laboratory), Alfanso García-Piñeres (immunologist), Sandra Silva (head microbiologist, HPV diagnostics laboratory), Ivannia Atmella (microbiologist, immunology laboratory), Margarita Ramírez (microbiologist, immunology laboratory). United States National Cancer Institute, Bethesda, MD Allan Hildesheim (coprincipal investigator and National Cancer Institute [NCI] coproject officer), Aimée R. Kreimer (co-investigator), Douglas R. Lowy (HPV virologist), Nora Macklin (trial coordinator), Mark Schiffman (medical monitor and NCI co-project officer), John T. Schiller (HPV virologist), Mark Sherman (QC pathologist), Diane Solomon (medical monitor and QC pathologist), Sholom Wacholder (statistician). SAIC, NCI-Frederick, Frederick, MD Ligia Pinto (head, HPV immunology laboratory), Troy Kemp (immunologist). Women s and Infants Hospital, Providence, RI Claire Eklund (QC cytology), Martha Hutchinson (QC cytology). Georgetown University, Washington, DC Mary Sidawy (histopathologist), DDL Diagnostic Laboratory, Netherlands Wim Quint (virologist, HPV DNA testing), Leen-Jan van Doorn (HPV DNA testing). Financial support. This work was supported by the NCI (contract N01- CP-11005), with funding support from the National Institutes of Health Office of Research on Women s Health. The Costa Rica HPV Vaccine Trial is a long-standing collaboration between investigators in Costa Rica and the NCI. Potential conflict of interest. For the NCI researchers Krystle A. Lang Kuhs, Mark Schiffman, Douglas Lowy, Mahboobeh Safaeian, John Schiller, Sholom Wacholder, Allan Hildesheim, and Aimée R. Kreimer: GlaxoSmithKline Biologicals (GSK) provided vaccine and support for aspects of the trial associated with regulatory submission needs of the company under a Clinical Trials Agreement (FDA BB-IND 7920) during the 4- year, randomized blinded phase of our study. The NCI and Costa Rica investigators are responsible for the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation of the manuscript. On efficacy-related manuscripts, GSK has the right to review and comment. GlaxoSmithKline Biologicals SA was provided the opportunity to review a preliminary version of this manuscript for factual accuracy, but the authors are solely responsible for final content and interpretation. The authors received no financial support or other form of compensation related to the development of the manuscript. Additional specific disclosure statement for Douglas Lowy: As part of my US government supported research at the National Cancer Institute (NCI)/National Institutes of Health (NIH), I am an inventor of technology that underlies the L1-based prophylactic virus-like particle (VLP) HPV vaccine and technology that underlies an L2-based candidate prophylactic HPV vaccine. The NIH has licensed the technology for the L1 VLP vaccine to Merck, the manufacturer of Gardasil, to GlaxoSmithKline, the manufacturer of Cervarix, and to Indian Immunologicals Ltd. The L2-based vaccine technology is the subject of a cooperative research and development agreement between the NCI, Johns Hopkins University, and Shantha Biotech, and has been licensed to Shantha, PanVax, Acambis Inc, and GSK. US federal law entitles me to a limited share of the royalties the NIH receives for these technologies. John Schiller: I am an inventor on US government-owned patents of the HPV VLP vaccine technology. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed. References 1. Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, CA Cancer J Clin 2007; 57: Jones RW, Baranyai J, Stables S. Trends in squamous cell carcinoma of the vulva: the influence of vulvar intraepithelial neoplasia. Obstet Gynecol 1997; 90: Jemal A, Simard EP, Dorell C, et al. Annual Report to the Nation on the Status of Cancer, , featuring the burden and trends in human papillomavirus (HPV) associated cancers and HPV vaccination coverage levels. J Natl Cancer Inst 2013; 105: Sturgeon SR, Brinton LA, Devesa SS, Kurman RJ. In situ and invasive vulvar cancer incidence trends (1973 to 1987). Am J Obstet Gynecol 1992; 166: Iversen T, Tretli S. Intraepithelial and invasive squamous cell neoplasia of the vulva: trends in incidence, recurrence, and survival rate in Norway. Obstet Gynecol 1998; 91: Messing MJ, Gallup DG. Carcinoma of the vulva in young women. Obstet Gynecol 1995; 86: Forman D, de Martel C, Lacey CJ, et al. Global burden of human papillomavirus and related diseases. Vaccine 2012; 30(suppl 5):F Arbyn M, de Sanjose S, Saraiya M, et al. EUROGIN 2011 roadmap on prevention and treatment of HPV-related disease. Int J Cancer 2012; 131: Paavonen J, Naud P, Salmeron J, et al. Efficacy of human papillomavirus (HPV)-16/18 AS04-adjuvanted vaccine against cervical infection and 1898 JID 2014:210 (15 December) Lang Kuhs et al

10 precancer caused by oncogenic HPV types (PATRICIA): final analysis of a double-blind, randomised study in young women. Lancet 2009; 374: Garland SM, Hernandez-Avila M, Wheeler CM, et al. Quadrivalent vaccine against human papillomavirus to prevent anogenital diseases. N Engl J Med 2007; 356: Garland SM. Efficacy of the HPV-16/18 AS04-adjuvanted vaccine against vulvar/vaginal intraepithelial neoplasia. In: IPV Conference, Berlin, Winer RL, Hughes JP, Feng Q, O Reilly S, Kiviat NB, Koutsky LA. Comparison of incident cervical and vulvar/vaginal human papillomavirus infections in newly sexually active young women. J Infect Dis 2009; 199: Garland SM, Insinga RP, Sings HL, Haupt RM, Joura EA. Human papillomavirus infections and vulvar disease development. Cancer Epidemiol Biomarkers Prev 2009; 18: Abalos AT, Harris RB, Nyitray AG, et al. Human papillomavirus type distribution among heterosexual couples. J Low Genit Tract Dis 2012; 16: Barzon L, Militello V, Pagni S, et al. Distribution of human papillomavirus types in the anogenital tract of females and males. J Med Virol 2010; 82: Rumbold AR, Tan SE, Condon JR, et al. Investigating a cluster of vulvar cancer in young women: a cross-sectional study of genital human papillomavirus prevalence. BMC Infect Dis 2012; 12: Howell-Jones R, de Silva N, Akpan M, et al. Prevalence of human papillomavirus (HPV) infections in sexually active adolescents and young women in England, prior to widespread HPV immunisation. Vaccine 2012; 30: Herrero R, Hildesheim A, Rodriguez AC, et al. Rationale and design of a community-based double-blind randomized clinical trial of an HPV 16 and 18 vaccine in Guanacaste, Costa Rica. Vaccine 2008; 26: Kreimer AR, Gonzalez P, Katki HA, et al. Efficacy of a bivalent HPV 16/ 18 vaccine against anal HPV 16/18 infection among young women: a nested analysis within the Costa Rica Vaccine Trial. Lancet Oncol 2011; 12: Herrero R, Wacholder S, Rodriguez AC, et al. Prevention of persistent human papillomavirus infection by an HPV16/18 vaccine: a communitybased randomized clinical trial in Guanacaste, Costa Rica. Cancer Discov 2011; 1: Kleter B, van Doorn LJ, ter Schegget J, et al. Novel short-fragment PCR assay for highly sensitive broad-spectrum detection of anogenital human papillomaviruses. Am J Pathol 1998; 153: Kleter B, van Doorn LJ, Schrauwen L, et al. Development and clinical evaluation of a highly sensitive PCR-reverse hybridization line probe assay for detection and identification of anogenital human papillomavirus. J Clin Microbiol 1999; 37: van Doorn LJ, Molijn A, Kleter B, Quint W, Colau B. Highly effective detection of human papillomavirus 16 and 18 DNA by a testing algorithm combining broad-spectrum and type-specific PCR. J Clin Microbiol 2006; 44: Herrero R, Quint W, Hildesheim A, et al. Reduced prevalence of oral human papillomavirus (HPV) 4 years after bivalent HPV vaccination in a randomized clinical trial in Costa Rica. PLOS One 2013; 8: e Rothman KJ, Boice JD. Epidemiologic analysis with a programmable calculator. new ed. Boston, MA, USA: Epidemiology Resources Inc, Agresti A. Categorical data analysis. 2nd ed. Hoboken, NJ: John Wiley and Sons Inc., Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986; 42: Lehtinen M, Paavonen J, Wheeler CM, et al. Overall efficacy of HPV-16/ 18 AS04-adjuvanted vaccine against grade 3 or greater cervical intraepithelial neoplasia: 4-year end-of-study analysis of the randomised, double-blind PATRICIA trial. Lancet Oncol 2012; 13: Hildesheim A, Herrero R, Wacholder S, et al. Effect of human papillomavirus 16/18 L1 viruslike particle vaccine among young women with preexisting infection: a randomized trial. JAMA 2007; 298: Reduced Vulvar HPV16/18 With Vaccination JID 2014:210 (15 December) 1899