Human papillomavirus vaccination and anal high-grade precancerous lesions and cancer—a real-world effectiveness study

Abstract

Background

Human papillomavirus (HPV) vaccination has shown high efficacy against anal HPV infection and lesions in clinical trials, and the HPV prevalence and type distribution in anal precancers and cancer predict a high preventable potential for HPV vaccination. However, the real-world effectiveness of HPV vaccination against anal high-grade lesions and cancer is yet to be shown.

Methods

We investigated HPV vaccine effectiveness against anal high-grade squamous intraepithelial lesion (HSIL) or worse in a nationwide cohort including all Danish women aged 17-32 years during October 2006 to December 2021 (n = 968 881). HPV vaccinations and first occurrence of anal HSIL or worse were retrieved from nationwide registries. Women were considered vaccinated after first dose and classified by age at vaccination. Using Cox regression, hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated for anal HSIL or worse according to vaccination status.

Results

During follow-up, the number of incident histological anal HSIL or worse cases was 37 in unvaccinated women, and less than 5 and 26 in women vaccinated at ages younger than 17 years and 17-32 years, respectively. The overall number of cancers was less than 5. Compared with unvaccinated women, the risk of histological anal HSIL or worse was reduced for women vaccinated at age younger than 17 years (HR = 0.30, 95% CI = 0.10 to 0.87). For women vaccinated at age 17-32 years, the hazard rate of anal HSIL or worse was 1.21 (95% CI = 0.73 to 2.03).

Conclusion

This is the first study to demonstrate that HPV vaccination at a younger age is associated with substantially reduced risk of anal HSIL or worse in the general population.

Human papillomavirus (HPV) is a necessary cause of virtually all cervical cancers, the vast majority of anal cancers, and varying but substantial proportions of other genital and oropharyngeal cancers (1,2). In a global perspective, an estimated 630 000 incident cancer cases in 2012 were attributable to HPV infection, and of these, 35 000 were of anal origin (2). HPV16 is by far the most carcinogenic HPV type in anal cancer and its precursor lesion anal high-grade squamous intraepithelial lesion (HSIL) (3,4). A recent large meta-analysis reported on the HPV type distribution in anal lesions by HIV status, which is strongly associated with anal cancer and precancer (3). In HIV-negative cancers, the estimated preventable proportions of cases was approximately 86%-92% and 95%-98% with the quadrivalent (HPV6, -11, -16, -18) and the nonavalent (HPV6, -11, -16, -18, -31, -33, -45, -52, -58) HPV vaccines, respectively, and with lower proportions in HIV-positive anal cancer patients (56%-77% and 75%-92%) (3). In anal HSIL, approximately 60% of cases were attributable to HPV types in the quadrivalent HPV vaccine, and with expected high gain in protection offered by the nonavalent vaccine as the prevalence of the 5 high-risk HPV types added in the nonavalent vaccine was high (30% and 38% in HIV-negative and HIV-positive women, respectively, and 45% and 57% in HIV-negative and HIV-positive men, respectively) (3).

For the bivalent (HPV16 and 18), quadrivalent, and nonavalent HPV vaccines, data from clinical trials have shown high efficacy against HPV-associated disease (5), and an increasing number of observational studies have also published data confirming vaccine effectiveness in a real-world setting. Initially, outcomes were genital warts (6) and cervical precancerous lesions (7), but the increasing length of follow-up time since licensure of the first HPV vaccine (2006) enables real-world studies assessing HPV vaccine effectiveness against outcomes with longer latency such as invasive cervical cancer (8-10) or rare outcomes such as vulvovaginal high-grade lesions (11). To our knowledge, data on the real-world effectiveness of HPV vaccination against anal high-grade lesions and cancer in the general population have not yet been published.

Denmark is a unique place to conduct follow-up studies and thereby effectiveness studies of HPV vaccination. The HPV vaccine was included in the Danish childhood vaccination program to girls aged 12 years in 2009 with simultaneous catch-up vaccination of older birth cohorts. In Denmark, we have nationwide individual-level and nearly complete registration of vaccination status and pathology diagnoses performed at all Danish pathology departments, and linkage is possible because of our unique personal identification number assigned to all Danish citizens.

In this setting, the aim of the present paper was to investigate HPV vaccine effectiveness against anal HSIL or worse among young Danish women.

Methods

Study population and follow-up

We conducted a nationwide cohort study using data from high-quality Danish registries holding information on HPV vaccination and redeemed prescription for other medication (12,13), pathology diagnoses (14), inpatient and outpatient hospital contacts (15), and education (16). The cohort was identified in the Danish Civil Registration System (17) and consisted of all Danish women aged 17-32 years during October 2006 to December 2021 and with no prior history of anal HSIL or worse. Linkage between the registries was performed by means of the personal identification number (17). Use of registry data for this study was approved by the Danish National Board of Health Data (FSEID-0045 and FSEID-4711).

HPV vaccination in Denmark and exposure assessment

In January 2009, the quadrivalent HPV vaccine was introduced free of charge to girls aged 12 years as part of the Danish childhood vaccination program and with catch-up vaccination for girls aged 13-15 years (birth cohorts 1993-1995) from October 2008 and for young women up to age 27 years (birth cohorts 1985-1992) in August 2012. Only recently (July 2019), the vaccination program was extended to boys. The vaccination program offered the bivalent HPV vaccine between February 2016 and October 2017, and hereafter the nonavalent vaccine has been used. Coverage of at least 1 dose has been high (≥80%) in both routinely vaccinated and catch-up cohorts (18,19). Denmark experienced a decline in vaccine uptake during 2015-2016 but with recovery to the uptake level before the decline (20).

We linked the study cohort to the Danish National Health Service Register (12) to retrieve information on HPV vaccination given in the free-of-charge vaccination program. The Health Service Register contains complete information of tax-funded public health-care activities delivered in primary care. Girls and women who were not covered by the vaccination program and had financed the vaccine themselves were identified in the Danish National Prescription Registry (13) holding individual-level data on prescriptions redeemed by Danish residents at all pharmacies in Denmark. Women were considered vaccinated after 1 dose of any of the 3 commercially available HPV vaccines.

Assessment of anal HSIL or worse

The cohort was followed for first diagnosis of anal HSIL or worse in the Danish Pathology Registry holding information on all cytologic and histologic diagnoses performed at pathology departments in Denmark since 1997 (14). The Pathology Registry uses the Systematized Nomenclature of Medicine based on codes for topography and morphology. The anal lesions were identified by the Systematized Nomenclature of Medicine topography codes T69. Because of the varying coding of anal precancerous lesions during our study period, anal HSIL or worse in our study refers to morphology codes for moderate and severe dysplasia, carcinoma in situ, HSIL (since June 2018), and squamous cell carcinomas.

Statistical analysis

The women entered the cohort on October 2006 or on their 17th birthday, whichever came last, and were followed until first occurrence of anal HSIL or worse, death, emigration, or end of follow-up (December 2021). The study period spanned 15 years meaning that the maximum attained age of women who were aged 17 years at cohort entry in 2006 was 32 years, and for that reason, all women were censored at their 32nd birthday. HPV vaccination was treated as a time-varying exposure, and the women could thus contribute unvaccinated and vaccinated person-time. We classified the vaccinated women according to age at vaccination. Initially, age at vaccination was assessed in 3 categories (younger than 17, 17-19, and 20-32 years) but subsequently dichotomized (younger than 17 and 17-32 years) as there were no outcomes registered among women vaccinated at ages 17-19 years.

The cumulative incidence of anal HSIL or worse was estimated according to age at vaccination and attained age. Hazard ratios (HRs) and corresponding 95% confidence intervals (CIs) for anal HSIL or worse according to vaccination status were estimated using Cox regression with age as underlying timescale. There was no evidence of deviations from the assumption of proportional hazards. Potentially confounding factors were selected a priori and included attained age (by use of age as underlying timescale) and maximum level of own, mother’s, or father’s education (short, medium, or long) derived from Statistics Denmark (16).

We performed 3 sensitivity analyses. First, we estimated vaccine effectiveness against anal HSIL by repeating the main analysis but with censoring of women with anal cancer at time of diagnosis. Secondly, we addressed the influence of previous high-grade cervical lesions as a surrogate marker for past or prevalent anal HPV infection at time of vaccination. In this analysis, women with cervical intraepithelial neoplasia grade 2 (CIN2) or worse prior to cohort entry were excluded. In addition, we adjusted the analysis for history of genital warts defined as the diagnoses (International Classification of Diseases version 10 (A63.0; version 8: 099.90) in the Patient Registry (15) or redemption of podophyllotoxin or imiquimod in the Prescription Registry (13).

All analyses were performed using the statistical software R version 4.2.2 (21).

Results

The cohort included 968 881 women contributing 7 290 830 person-years. At cohort entry, 374 489 (38.7%) women were vaccinated at age younger than 17 years, and 594 392 (61.3%) were unvaccinated (Table 1). During follow-up, additional 198 214 (20.5%) women were vaccinated at ages 17-32 years, and the remaining 396 178 (40.9%) women remained unvaccinated. Among the vaccinated women, 71.8% received 3 doses, 21.4% received 2 doses, and 6.8% received 1 dose. Altogether, 37 incident cases of anal HSIL or worse were diagnosed among unvaccinated women. In vaccinated women, the corresponding numbers were less than 5 and 26 for those aged younger than 17 years and 17-32 years at first vaccine dose, respectively. The number of anal cancers constituted less than 5 of the total number of outcomes. All cases of anal HSIL or worse were based on histology.

Figure 1 shows the cumulative incidence of histological anal HSIL or worse by age at vaccination and attained age. In the unvaccinated cohort and among women vaccinated at age 17-32 years, the incidence showed the same pattern with a rise starting in the early 20s and reaching a maximum of 0.02% at age 32 years. In women vaccinated at age younger than 17 years, the incidence of anal HSIL or worse remained below 0.002% with increasing attained age.

The overall hazard ratio of histological anal HSIL or worse for the comparison of vaccinated women with unvaccinated women was 0.59 (95% CI = 0.35 to 0.99), and after additional adjustment for educational level, the hazard ratio was 0.63 (95% CI = 0.37 to 1.06) (data not shown in table). After stratification according to age at vaccination, the hazard ratio of anal HSIL or worse was statistically significantly reduced for women vaccinated at age younger than 17 years (HR = 0.30, 95% CI = 0.10 to 0.87); for women vaccinated at 17-32 years, the hazard ratio of anal HSIL or worse was 1.21 (95% CI = 0.73 to 2.03).

In the analysis exclusively of precancerous lesions, the hazard ratio of histological anal HSIL was 0.30 (95% CI = 0.10 to 0.90) among women vaccinated at age younger than 17 years and 1.27 (95% CI = 0.75 to 2.15) among women vaccinated at age 17-32 years. Moreover, hazard ratios of anal HSIL or worse were similar to those in the main analysis after exclusion of women with CIN2 or worse prior to cohort entry or after additional adjustment for a history of genital warts (data not shown).

Discussion

To our knowledge, this is the first study to present real-world data on HPV vaccine effectiveness against anal HSIL or worse. Estimates for the women vaccinated at the youngest age showed that these women had a 70% lower hazard of developing histologically diagnosed anal HSIL or worse compared with women not vaccinated. Nearly all outcomes corresponded to anal HSIL, and the majority of vaccinated women in our study population had received the quadrivalent vaccine. For these reasons, our data primarily reflect vaccine effectiveness against anal HSIL provided by protection against HPV16 and 18. Together with previous data on cervical (8-10) and vulvovaginal (11) high-grade lesions and cancer, there is now substantial real-world results supporting that the HPV vaccine is effective in preventing cervical cancer and high-grade lesions on all female anogenital sites.

Anal cancer is rare but with increasing incidence reported in many populations all over the world, including Denmark (22,23). In the general population, the incidence is higher in women than in men in most countries (22,23). However, anal cancer is particularly frequent among HIV-positive individuals, men who have sex with men (MSM), women with other HPV-associated gynecological diseases, and people with iatrogenic immunosuppression such as solid organ transplant recipients (24). MSM living with HIV are by far the group with the highest risk with an incidence rate of 85 per 100 000 person-years according to a recent meta-analysis (24). In all high-risk groups, age is a strong risk factor for the development of anal cancer (24).

Even though our study population included young women from the general Danish female population with a low incidence of anal cancer and precancer (23), we already now, approximately 13 years after licensure of the HPV vaccine, observed high HPV vaccine effectiveness against histological anal HSIL or worse. No anal HSIL or worse outcomes in our study were based on cytology, which may be used for screening purposes (25). Most of the vaccinated women in our study received the quadrivalent HPV vaccine as the bivalent was only used for approximately 1.5 years, and the nonavalent vaccine was first included in the vaccination program at the end of 2017. Our results therefore primarily reflect protection against HPV16 and 18. Moreover, the number of cancer outcomes were limited, and results from the sensitivity analysis of exclusively HSIL were similar to those in the main analysis indicating that the observed vaccine effectiveness of 70% applies primarily to anal HSIL, which is in accordance with the reported prevalence of HPV16 and 18 in high-grade anal lesions (3).

Data from randomized controlled clinical trials have shown high efficacy of the HPV vaccine against anal HPV16 and 18 infections (26,27). A nested analysis within the Costa Rica Vaccine Trial including 4210 healthy women randomly assigned at ages 18-25 years showed that efficacy of the bivalent HPV vaccine against anal HPV16 and 18 infection, measured 4 years postvaccination, was 62.0% (95% CI = 47.1% to 73.1%) overall and 83.6% (95% CI = 66.7% to 92.8%) in a restricted analysis of women being HPV naïve at enrollment (cervical HPV16 and 18 negative and HPV16 and 18 seronegative) (26). The preventive potential of the quadrivalent HPV vaccine against histologically verified anal high-grade lesions in MSM has also been documented (27-29). Palefsky et al. (27) included 602 young, HIV-negative MSM and reported that vaccine efficacy against anal HSIL was 74.9% (95% CI = 8.8% to 95.4%) in the per protocol analysis and 54.2% (95% CI = 18.0% to 75.3%) in the intention-to-treat population. In 144 enrolled MSM living with HIV, no incident cases of anal HSIL associated with HPV types included in the quadrivalent vaccine were detected at 2 years postvaccination follow-up (29).

Our study population only included women, as boys were not included in the Danish HPV vaccination program until 2019. Several other countries have also introduced gender-neutral HPV vaccination (30). We anticipate that the observed vaccine effectiveness would be similar in the Danish male population as the proportions of anal cancers attributable to the HPV types included in the quadrivalent and the nonavalent HPV vaccines, respectively, are similar in men and women (3). Moreover, vaccine efficacy against anal high-risk HPV infection seems to be similar by sex (26,27).

Our results emphasize the importance of vaccinating at an early age. Because of the registry-based design of our study, we had no information on sexual behavior. However, a large questionnaire study of 20 478 Danish girls showed that median age of sexual debut is 16 years (31), and this was the rationale for our categorization of age at vaccination. In our study, the risk of anal HSIL or worse was not reduced among women vaccinated at age 17-32 years. We have previously observed a similar pattern regarding the risk of cervical cancer (8) and severe vulvovaginal lesions (11). In this group of women, some paid for opportunistic HPV vaccination, and it is conceivable that our findings to some extent may be biased by self-selection. Moreover, we had no information on prevalent HPV infection at time of vaccination, and length of postvaccination follow-up time for these women was likely too short to observe the true benefit of HPV vaccination against anal lesions (32). For those people who already have an anal high-risk HPV infection, however, secondary prevention is crucial. Screening against anal cancer in high-risk populations is investigated using methods known from cervical cancer screening (25), but introduction of formal screening programs of high-risk populations has awaited recently published evidence on the risk reduction of anal cancer following treatment of anal HSIL (33). In that study, which is a phase III randomized clinical trial, 4446 persons living with HIV and with anal HSIL received either HSIL treatment or active monitoring, and the results showed that the rate of progression to anal cancer was 57% lower in the treatment group compared with the group that was actively monitored (33).

The main strength of our study is the nationwide design with virtually complete individual-level information on HPV vaccination status and follow-up with registration of all histologically verified cases of anal HSIL or worse. There are also limitations that must be acknowledged. Because of the real-world study design, some vaccinated women may have had a past or prevalent anal HPV infection at time of vaccination. To address this issue, we stratified the analyses according to age at vaccination and added 2 sensitivity analyses with exclusion of women with cervical lesions corresponding to CIN2 or worse or with adjustment for history of genital warts, respectively. Finally, there may also be some selection bias regarding vaccination uptake (34), and this was handled by adjusting the analyses for educational level as a marker of socioeconomic status and a proxy for lifestyle factors. Our study may, however, still be subject to residual confounding.

In conclusion, this early report on HPV vaccine effectiveness against histological anal HSIL or worse indicates substantially reduced risk among women vaccinated at a younger age. With these results, there is now evidence on efficacy from clinical trials and on real-world effectiveness that the HPV vaccination is effective in preventing high-grade lesions on all female anogenital sites. Our results may also suggest great potential benefits of HPV vaccination in men.

Data availability

The data analyzed in this study are pseudo-anonymized data available from Statistic Denmark under license for this study. The data are available after application to and permission from Statistic Denmark.

Author contributions

Louise Baandrup, PhD, MD (Conceptualization; Investigation; Methodology; Writing—original draft), Thomas Maltesen, PhD (Formal analysis; Investigation; Methodology; Validation; Writing—review & editing), Christian Dehlendorff, PhD (Conceptualization; Investigation; Methodology; Validation; Writing—review & editing), and Susanne K. Kjaer, DMSc (Conceptualization; Investigation; Methodology; Resources; Writing—original draft; Writing – review & editing).

Funding

This work was supported by the Mermaid project [Mermaid 2].

Conflicts of interest

SKK has a research grant through her research institution from Merck. LB, TM, and CD have declared no conflicts of interest.

Acknowledgements

The funder did not play a role in the design of the study; the collection, analysis, and interpretation of the data; the writing of the manuscript; and the decision to submit the manuscript for publication. The authors are grateful to Rasmus Hertzum-Larsen for data management assistance.

References