aA total of 3013 neither initiated human papillomavirus (HPV) vaccination nor had a missed opportunity for HPV vaccination from 11 through 20 years of age.
bScreened following initiation of HPV vaccination or, if unvaccinated, the first missed opportunity for HPV vaccination after January 1, 2007.
Time (years) to the first abnormal cervical cytology result after either initiation of HPV vaccination or the first missed vaccination opportunity, if unvaccinated, in the entire study population (n = 2828) (A) and among those aged 11 through 14 years (n = 124) (B).
Customize your JAMA Network experience by selecting one or more topics from the list below.
Hofstetter AM, Ompad DC, Stockwell MS, Rosenthal SL, Soren K. Human Papillomavirus Vaccination and Cervical Cytology Outcomes Among Urban Low-Income Minority Females. JAMA Pediatr. 2016;170(5):445–452. doi:10.1001/jamapediatrics.2015.3926
The quadrivalent human papillomavirus (HPV) vaccine was licensed for use in 9- through 26-year-old females in 2006. Postlicensure studies in Australia, Denmark, and Canada have demonstrated vaccine effectiveness against abnormal cervical cytology results. However, there are limited data describing postlicensure effectiveness in the United States, particularly among minority females at higher risk for HPV infection and cervical cancer.
To examine the effect of HPV vaccination on abnormal cervical cytology results among minority females.
Design, Setting, and Participants
Retrospective cohort study conducted between January 2007 and January 2014 at 16 academically affiliated community clinics serving a low-income minority population. Included in this study was a population-based sample of 16 266 females aged 11 through 20 years as of January 1, 2007, who received care at a participating clinic on or after that date.
Human papillomavirus vaccination, stratified by the number of doses.
Main Outcomes and Measures
Cervical cytology abnormality following either HPV vaccination or, if unvaccinated, the first missed opportunity for HPV vaccination after January 1, 2007. Abnormalities were defined as atypical glandular cells, atypical squamous cells of undetermined significance, atypical squamous cells, cannot exclude a high-grade squamous intraepithelial lesion, low-grade squamous intraepithelial lesions, or high-grade squamous intraepithelial lesions.
There were 4127 female patients who initiated quadrivalent HPV vaccination or had their first missed HPV vaccination opportunity from 11 through 20 years of age and underwent subsequent cervical cytology screening. The patients were primarily Spanish speaking (n = 2297; 58.3%) and publicly insured (n = 3801; 92.1%). The detection rate for an abnormal cervical cytology result during the observation period was lower among vaccinated (≥1 dose) (79.1 per 1000 person-years) vs unvaccinated (125.7 per 1000 person-years) females. The risk for an abnormal cervical cytology result was lower among vaccinated vs unvaccinated females (hazard ratio [HR], 0.64; 95% CI, 0.57-0.73), particularly if the 3-dose series was completed (HR, 0.48; 95% CI, 0.41-0.56) or if the vaccine was administered from 11 through 14 years of age (≥1 dose: HR, 0.36; 95% CI, 0.16-0.79; 3 doses: HR, 0.27; 95% CI, 0.12-0.63). This protective effect remained after adjusting for demographics, clinic type, abnormal baseline cervical cytology result, and baseline Chlamydia screening (as proxy for sexual experience).
Conclusions and Relevance
This study demonstrated the HPV vaccine is effective in a real-world setting of high-risk patients with variable HPV vaccination patterns.
Human papillomavirus (HPV) is the most prevalent sexually transmitted infection in the United States.1 Persistent HPV infection can cause genital warts, anogenital cancers, and certain orophargyneal cancers, leading to significant morbidity and mortality.2 Data suggest that low-income minority females may be disproportionately affected.3-7
In June 2006, the quadrivalent HPV vaccine was licensed for use among females aged 9 through 26 years in the United States.8 The Advisory Committee on Immunization Practices subsequently recommended HPV vaccination as a 3-dose series for all females aged 11 through 26 years.9 End-of-study per-protocol analyses from 3 clinical trials demonstrated a 98.2% efficacy against HPV 6–, 11–, 16–, and 18–related CIN 2+ lesions.10 Postlicensure studies have shown a positive impact on HPV prevalence and genital lesions at the individual and population levels.11-16 Recent data also have demonstrated vaccine effectiveness against abnormal cervical cytology results.17-22 For example, population-based linkage studies from Australia, Denmark, and Canada revealed a lower risk for cervical abnormalities among HPV-vaccinated vs unvaccinated females.17-19,21,22
Examination of HPV vaccine effectiveness among minority populations, individuals engaging in sexual risk behaviors, and those with variable HPV vaccination patterns is warranted. For example, only 30% of participants in the HPV vaccine trials represented a minority race/ethnicity.10 A study of sexually active low-income minority females (n = 235) found a lower prevalence of abnormal cervical cytology results among vaccinated vs unvaccinated individuals,20 yet further investigation in a larger sample is needed. Similarly, clinical trials excluded individuals with prior confirmed HPV disease and/or greater than 4 lifetime partners, and per-protocol analyses excluded participants who at 1 month after the third dose were polymerase chain reaction positive for HPV vaccine types that they were naive to at enrollment or who failed to follow the protocol, including receipt of 3 doses within 12 months.10 Improved understanding of the vaccine’s protective effect in real-world settings is needed given that many initiate HPV vaccination after sexual debut23 and often fail to complete the 3-dose series,24 even within 12 months of initiation.25
In the present study, the risk for abnormal cervical cytology results was assessed among vaccinated and unvaccinated low-income minority females from an underserved urban community.
Question What are the cervical cytology outcomes following human papillomavirus (HPV) vaccination in a cohort of urban low-income minority females?
Findings The risk for an abnormal cervical cytology result was 36% lower among HPV-vaccinated compared with unvaccinated individuals. The risk for an abnormal cervical cytology result was 73% lower among young adolescents who were fully vaccinated compared with those who remained unvaccinated.
Meaning A protective effect was also observed among those with partial vaccination (2 doses) and those who initiated HPV vaccination in later adolescence.
This retrospective cohort study was conducted in 4 pediatric clinics, 6 school-based health clinics (SBHCs), 1 family medicine clinic, 1 family planning clinic, and 4 obstetric/gynecology clinics affiliated with a large academic medical center serving a low-income minority population in New York City. The quadrivalent HPV vaccine has been routinely given to female patients at these clinics since January 2007. The cervical cytology screening protocol changed over the study period. Prior to the American Congress of Obstetricians and Gynecologists issuing new screening recommendations in December 2009,26 the clinics typically initiated routine screening around 3 years after sexual debut, resulting in approximately one-fifth of patients undergoing their first screening between 13 and 20 years of age.27 In response to the new American Congress of Obstetricians and Gynecologists guidelines, these sites deferred routine screening until age 21 years, regardless of age at initiation of sexual activity, resulting in only 6% undergoing their first screening between 13 and 20 years of age.27 Screening patterns also differed by clinic setting (ie, more screening at obstetric/gynecology and family-planning clinics and less screening at SBHCs compared with pediatric clinics).
Adolescent and young adult females aged 11 through 20 years as of January 1, 2007, with at least 1 visit to a participating clinic on or after that date were eligible for study participation. Additional inclusion criteria were (1) initiation of HPV vaccination or, if unvaccinated, a missed opportunity for HPV vaccination from 11 through 20 years of age and (2) cervical cytology screening during the observation period. The study age range was selected a priori given the focus on HPV vaccination during early adolescence, yet recognizing that some patients may not receive their first dose until older adolescence or beyond. This study was approved by the Columbia University Medical Center Institutional Review Board, with a waiver of patient consent for a minimal-risk study.
Demographic and clinic visit data were obtained from the medical center’s registration system. Cervical cytology and Chlamydia screening data through January 10, 2014, were collected from the medical center’s clinical data warehouse. Human papillomavirus vaccine data through January 10, 2014, were obtained from the medical center’s immunization registry, EzVac. EzVac includes all vaccine doses administered to patients at the hospital and affiliated clinics. It also synchronizes with the New York Citywide Immunization Registry. Facilities are required to report to the Citywide Immunization Registry all vaccine administrations to patients younger than 19 years in New York City and an estimated 94% regularly do so.28 Thus, data in the present study also include HPV vaccine doses administered at nonstudy sites in New York City.
The primary outcome measure of interest was an abnormal cervical cytology result. Abnormalities were defined as atypical squamous cells of undetermined significance, atypical glandular cells, atypical squamous cells cannot exclude a high-grade squamous intraepithelial lesion, low-grade squamous intraepithelial lesion, and high-grade squamous intraepithelial lesion.29 A secondary outcome measure of interest was a high-grade abnormal cervical cytology result, defined as atypical glandular cells, atypical squamous cells cannot exclude a high-grade squamous intraepithelial lesion, and high-grade squamous intraepithelial lesion. The observation period began on the day following HPV vaccination initiation for vaccinated individuals or following the first missed HPV vaccination opportunity after January 1, 2007, for unvaccinated individuals. It continued until either the first abnormal cervical cytology result or, if all normal, the last cervical cytology screening. Screening in the 28-day postvaccination window was excluded from analysis of the given dose to address the lag time needed for the vaccine to take effect.18
Independent variables included the number of HPV vaccine doses; age at HPV vaccination initiation or the first missed HPV vaccination opportunity (11-14, 15-16, 17-18, and 19-20 years); age as of January 1, 2007; language (Spanish, English, or other); insurance (public, private, or uninsured); clinic type (obstetrics/gynecology, family planning, pediatric, family medicine, or SBHC); abnormal baseline cervical cytology result; and baseline Chlamydia screening (as a proxy for sexual experience) before HPV vaccination initiation or the first missed HPV vaccination opportunity.
Demographic characteristics, screening history, and HPV vaccination status were described using frequency distributions. Detection rates were calculated as the number of events per 1000 person-years. Kaplan-Meier analysis was used to assess time to an abnormal cervical cytology result during the observation period. Individuals without an abnormality were censored at the time of their last normal screening. Cox proportional hazards regression was used to estimate hazard ratios (HRs) and associated 95% CIs, examining the association between vaccination status and abnormal cervical cytology result, adjusting for demographic characteristics (age, language, and insurance), clinic type, abnormal baseline cervical cytology result, and baseline Chlamydia screening. An adjusted HR (aHR) less than 1 signified a lower likelihood of an abnormal cervical cytology result. To assess the effect of partial vaccination, as well as the timing of vaccination, analyses were stratified by the number of HPV vaccine doses and age at HPV vaccination initiation or, if unvaccinated, the first missed HPV vaccination opportunity.
Of note, individuals who received more than 1 HPV vaccine dose but had no cervical cytology screening after the second dose (n = 180) and those who received more than 1 dose but had their first abnormality between the first and second doses (n = 70) were included in the 1-dose only analytic group. Similarly, individuals who received 3 doses but had no screening after the third dose (n = 176) and those who received 3 doses but had an abnormality between the second and third doses (n = 46) were included in the 2-dose only analytic group. All analyses were performed using SAS Version 9.3 (SAS Institute).
Of the 16 266 individuals fulfilling age and visit criteria, 13 253 initiated HPV vaccination or, if unvaccinated, had their first missed opportunity for HPV vaccination from 11 through 20 years of age (Figure 1). In total, 4127 underwent subsequent cervical cytology screening. The proportion with subsequent screening did not differ by vaccination status except among those aged 19 through 20 years, where fewer vaccinated compared with unvaccinated individuals had undergone screening (57.1% vs 61.8%; P < .05). The median observation period duration was 2.6 years (maximum, 7.0 years).
Individuals were predominantly Spanish speaking (n = 2297; 58.1%) and publically insured (n = 3801; n = 92.0%) (Table 1). They were distributed across clinic types, although most—particularly in the unvaccinated population—received care in reproductive health settings. More than half (60.5%) had initiated HPV vaccination. A higher proportion of patients in the pediatric, SBHC, and family medicine clinics were vaccinated (79.1%) compared with patients in the other clinics (55.0%) (P < .001). Fewer vaccinated individuals exhibited an abnormal baseline cervical cytology result (P < .001) and had undergone prior Chlamydia screening (as a proxy for sexual experience) (P < .001) compared with unvaccinated individuals. However, after stratifying by age group, abnormal baseline cervical cytology result differed by vaccination status only among those aged 17 through 18 years and baseline Chlamydia screening varied by vaccination status only in the older age groups (those aged 15-16, 17-18, and 19-20 years).
The detection rate for abnormal cervical cytology during the observation period among fully vaccinated individuals (3 doses) (58.2 per 1000 person-years) was less than half the rate observed among unvaccinated individuals (125.7 per 1000 person-years) (Table 2). Further, for those with abnormal cervical cytology, the duration to the first abnormality was longer among fully vaccinated vs unvaccinated individuals, as illustrated in the entire study population and more markedly in those aged 11 through 14 years (Figure 2).
In the unadjusted model, the risk for an abnormal cervical cytology result among vaccinated compared with unvaccinated females decreased with earlier age at vaccination (Table 2). A lower risk was also observed among those who received 2 or 3 doses prior to cytology screening compared with unvaccinated individuals; receipt of only 1 dose did not significantly reduce the risk. In the 2-dose only group (n = 604), the median (SD) duration between the first and second doses was 4 (9.4) months; 228 (37.7%) had at least 6 months between their first and second doses. Their risk of an abnormal cervical cytology result was similar to those who received 2 doses irrespective of second dose timing, except among those aged 17 through 18 years who received 2 doses separated by at least 6 months (n = 100) where a lower risk compared with unvaccinated individuals was also observed (HR, 0.65; 95% CI, 0.42-0.99).
In the adjusted model, including age (as of January 1, 2007), language, insurance, clinic type, abnormal baseline cervical cytology result, and Chlamydia screening, the risk for an abnormal cervical cytology result no longer differed between those aged 15 through 16 years who received 1 dose only vs no doses (aHR, 1.45; 95% CI, 0.88-2.37) or between those aged 17 through 18 years (aHR, 0.81; 95% CI, 0.64-1.01) and 19 through 20 years (aHR, 0.85; 95% CI, 0.68-1.05) who received 1 or more doses compared with no doses (Table 2).
There were 122 individuals (3.0%) with high-grade lesions. The detection rate for a high-grade lesion during the observation period among fully vaccinated females (4.2 per 1000 person-years) was one-third the rate observed among unvaccinated females (12.6 per 1000 person-years). Thus, the risk for a high-grade lesion was lower for fully vaccinated vs unvaccinated individuals (HR, 0.38; 95% CI, 0.20-0.57). There was no significant difference in the risk for a high-grade lesion between those who received 2 doses only compared with those who remained unvaccinated (HR, 0.82; 95% CI, 0.46-1.44).
This study demonstrates a strong protective effect of the quadrivalent HPV vaccine on cervical cytology outcomes among urban, low-income, minority patients with variable HPV vaccine administration between 2007 and 2014. The effect appeared to be greatest when vaccinating at a younger age and on completion of the 3-dose series, although a reduced risk for an abnormal cervical cytology result was apparent among older vaccinated individuals and those with partial vaccination. This study provides crucial information about the HPV vaccine in minority populations and females engaging in high-risk sexual behaviors who arguably were underrepresented in the HPV vaccine clinical trials and postlicensure studies, yet are at increased risk for cervical cancer. These real-world findings, which complement existing HPV vaccine efficacy/effectiveness data, could also have implications for the HPV vaccination recommendations with respect to timing and number of doses, including for the recently licensed nonavalent HPV vaccine.30
In this study, the overall risk for an abnormal cervical cytology result was lower among vaccinated compared with unvaccinated adolescent and young adult females. This is consistent with clinical trial findings and recent observational studies.10,17-19,21,22 In addition, the risk was lowest when HPV vaccination was initiated between 11 and 14 years of age. This finding is in accordance with existing data suggesting that younger adolescents mount a more robust response to HPV vaccination than older patients.31-33 It also likely demonstrates the greater protection afforded by vaccinating prior to sexual debut—previously shown to occur at a mean age of 15.7 years in this clinic population23—because HPV acquisition typically occurs shortly thereafter.34 These findings, coupled with our understanding of health care use patterns,35 lend support to existing Advisory Committee on Immunization Practices recommendations targeting HPV vaccination initiation among those aged 11 through 12 years.2
In this study, the risk for an abnormal cervical cytology result was lowest among those who completed the 3-dose series. However, a protective effect against any abnormality was also observed among those who were partially vaccinated, consistent with recent observational studies in Australia and Canada.17,19,22 Currently, the World Health Organization recommends a 2-dose regimen (0 and 6 months) for adolescent females younger than 15 years based on immunogenicity data and other pragmatic considerations33,36,37; this regimen has been implemented in many settings worldwide.38 Although the present study did not show a reduced risk for any abnormality among those aged 11 through 14 years who received 2 doses, it was likely underpowered to detect such an effect, particularly because only a few (12 of 25) of those aged 11 through 14 years received their second dose at least 6 months after the first dose.
This study was conducted in a high-risk population of urban, low-income, minority females, many of whom had undergone sexual health screening, a proxy for sexual experience, before the observation period. Earlier data from this clinic population revealed a lower mean age at sexual debut and higher rate of abnormal screening than has been shown nationally.23,39-41 This, along with the variable timing and number of HPV vaccine doses, certainly could have attenuated any benefit of HPV vaccination. Nonetheless, a reduced risk for abnormal cervical cytology results was still observed, lending support to the recommendation that all females, irrespective of prior sexual experience and screening results, complete the HPV vaccine series.
There were several limitations to this study. First, decreased cervical cytology screening after implementation of the 2009 American Congress of Obstetricians and Gynecologists guidelines likely altered detection rates over time and differentially across demographic groups and clinic settings.27 Similarly, abnormal baseline cervical cytology results and Chlamydia screening (as a proxy for sexual experience) differed by HPV vaccination status among older individuals. Although the protective effect of HPV vaccination remained significant after adjusting for these variables, other baseline differences between vaccinated and unvaccinated individuals (ie, with respect to risk behaviors) may exist that were not adjusted for in the analyses. In addition, there was inherent complexity with respect to HPV vaccination, the timing of cervical cytology screening, and the pattern of results. To address this, we excluded all subsequent screening results after the first abnormal cervical cytology result. Additionally, there could have been underreporting of HPV vaccination, resulting in misclassification. However, this is unlikely given routine vaccine data abstractions and reporting mandates. Similarly, cervical cytology screening performed at outside facilities would not have been captured here, although we do not anticipate this to differ by vaccination status. Further, although Chlamydia screening was used as a proxy for sexual experience, there is limited knowledge of actual HPV exposure or infection before HPV vaccination in the study population. Last, this study included a relatively modest number of adolescent and young adult females from 1 urban, low-income, minority community. Although our findings are consistent with population-based studies, further investigation in a larger cohort of minority females, including from other settings, would be beneficial.
This study demonstrated that the quadrivalent HPV vaccine was effective in a real-world setting of high-risk patients with variable HPV vaccination patterns. The greatest effect was observed among those who completed the series during early adolescence. A protective effect was also apparent among individuals vaccinated at an older age and was likely attenuated by residual confounders (ie, given suboptimal measurement of background prevalent infection). Additionally, a decreased risk was observed following 2 doses, despite few individuals receiving them at least 6 months apart. Longer follow-up in a larger sample is needed, particularly after more widespread uptake of the nonavalent HPV vaccine, which offers protection against 5 additional HPV genotypes.
Corresponding Author: Annika M. Hofstetter, MD, PhD, MPH, Division of General Pediatrics, Department of Pediatrics, University of Washington, 1900 9th Ave, Room 947, Seattle, WA 98101 (firstname.lastname@example.org).
Accepted for Publication: October 22, 2015.
Published Online: March 14, 2016. doi:10.1001/jamapediatrics.2015.3926.
Author Contributions: Dr Hofstetter had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Hofstetter, Soren.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Hofstetter.
Obtained funding: Soren.
Study supervision: Stockwell, Rosenthal, Soren.
Conflict of Interest Disclosures: Dr Hofstetter receives support for a separate investigator-initiated study from the Pfizer Medical Education Group. No other disclosures were reported.
Funding/Support: This study was supported in part by a research grant from the Merck Investigator-Initiated Studies Program.
Role of the Funder/Sponsor: Merck Sharp & Dohme Corp had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. The company did receive the manuscript prior to submission. The opinions expressed in this article are those of the authors and do not necessarily represent those of the Merck Sharp & Dohme Corp.
Previous Presentations: Information included in this article was presented at the International Papillomavirus Conference; August 24, 2014; Seattle, Washington; and the Pediatric Academic Societies Annual Meeting; April 26, 2015; San Diego, California.
Additional Contributions: We thank New York–Presbyterian Hospital for its support of the EzVac immunization registry and the New York–Presbyterian Hospital Ambulatory Care Network. We also thank Oscar Peña, JD (New York–Presbyterian Hospital), and Balendu DasGupta, MS (Columbia University), for their work on EzVac, and Karthik Natajaran, PhD (Columbia University), and Alla Babina, MS (Columbia University), for their assistance with data collection. None received compensation for their contributions.
Create a personal account or sign in to: