Trends in Human Papillomavirus-Associated Cancers, Demographic Characteristics, and Vaccinations in the US, 2001-2017

Abstract

Importance: Nearly 45 000 human papillomavirus (HPV)-associated cancers are diagnosed annually in the US. The HPV vaccine has been approved since 2006, but information on the association between vaccination and the incidence of HPV-attributable cancers is unclear.

Objective: To evaluate the potential association of screening and vaccination on the trends of HPV-associated cancers.

Design, setting, and participants: A retrospective, population-based cross-sectional study was conducted using data on HPV-associated (oropharyngeal squamous cell carcinoma [SCC], anal/rectal SCC, vulvar SCC, vaginal SCC, cervical carcinoma, and penile SCC) cancers from the US Cancer Statistics Public Use Database, representing 99% of the US population, between January 1, 2001, and December 31, 2017; HPV vaccination and screening data from the Behavioral Risk Factor Surveillance between January 1, 2001, and December 31, 2016; and TeenVaxView between January 1, 2008, and December 31, 2018. National Cancer Database and Behavioral Risk Factor Surveillance were used to correct for hysterectomy. Data analysis was performed from April 1, 2020, to June 30, 2021.

Exposures: Patient demographic characteristics, including age, race and ethnicity, sex, region, and vaccination status.

Main outcomes and measures: The main outcomes examined in this study were diagnoses of any HPV-associated cancer and HPV vaccination status.

Results: A total of 657 317 HPV-associated cancers (exact ages not collected by the United States Cancer Statistics); of these, 264 019 (40.2%) developed in men and 393 298 (59.8%) in women; 14 520 individuals (2.2%) were non-Hispanic Asian/Pacific Islander, 74 641 (11.4%) were non-Hispanic Black, 59 841 (9.1%) were Hispanic, and 499 899 were non-Hispanic White (76.1%). More than half (206 075 [52.4%]) of cancers in women were cervical, whereas most (211 421 [80.1%]) cancers in men were oropharyngeal. In female adolescents (aged 13-17 years), the vaccination rate increased from 37.2% to 69.9% from 2008 to 2018 (annual percent change: 6.57% [95% CI, 5.83%-7.32%]). Before vaccination approval, cervical cancer rates in the 20- to 24-year age group were decreasing at 2.29% annually (P = .045); after vaccine approval, this rate has been decreasing at 9.50% (P = .003). In men, annual increases were noted in oropharyngeal (2.71%) and anal/rectal (1.83%) cancers (P < .001); in women, the incidence of oropharyngeal remained stable but anal/rectal cancer increased at 2.83% every year (P < .001).

Conclusions and relevance: In the US, cervical cancer rates have decreased at a population level, especially in younger women. The findings of this study suggest this decrease may be associated with vaccination. Given the increase in oropharyngeal and anal/rectal cancers, particularly in men, it may be important to highlight vaccination uptake in both sexes.

Figure 1.. Average Annual Percent Change (AAPC) of Human Papillomavirus Virus (HPV)-Associated Cancers in US Cancer Statistics Public Use Databases From 2001 to 2017

Trends based on incidence were analyzed using the Joinpoint Regression Program, version 4.8.0.1, allowing up to 3 joinpoints. The AAPC is a summary measure of the trend over a prespecified fixed interval. It was computed as a weighted average of the APC from the joinpoint model, with the weights equal to the length of the APC interval. Vertical lines indicate the associated 95% CIs. SCC indicates squamous cell carcinoma. ^aThe AAPC is significantly different from 0 (P < .05).

Figure 2.. Age-Specified Incidences and Trends of Cervical Squamous Cell Carcinoma in Individuals Aged 20 to 24 Years and Human Papillomavirus Virus (HPV) Vaccination Rate (>1 Dose) in Adolescents Aged 13 to 17 Years

Trends based on incidence were analyzed using the Joinpoint Regression Program, version 4.8.0.1, allowing up to 3 joinpoints. The average annual percent change (AAPC) is a summary measure of the trend over a prespecified fixed interval. It was computed as a weighted average of the APC from the joinpoint model, with the weights equal to the length of the APC interval. Vertical lines indicate the associated 95% CIs.