Human Papillomavirus Vaccination and Screening in the Elimination of HPV-Associated Cancers: Evidence-Based Randomized Trials

Preventive medicine is largely about identification of causes of diseases and their removal. Cervical cancer in Finland is firstly a showcase and secondly a use-case of preventive medicine. Firstly, etiological studies on cervical cancer were for long confounded by the fact that sexually transmitted infections are surrogates of both risktaking behaviour in adolescent and young adult population, and occurrence of cervical cancer in middle-aged women. Identifying oncogenic human papillomaviruses (HPVs) as the true cause of cervical cancer among the multitude of different sexually transmitted micro-organisms required a Nobel-prize winning vision which was initially supported only by case-series evidence. It also required a paradigm shift that was facilitated by a correctly done epidemiological study and increased understanding on the molecular basis of exposure misclassification. All this was understood only after the etiological enigma had been resolved. Secondly, since the sexual revolution in 1960’s first facilitated increase in risk-taking sexual behaviour associated sexually transmitted infections’ incidence, and subsequently resulted in an increase in the incidence of cervical cancer. In the below Finnish use-case, the role of different causal (HPV16/18/31/45), intervening (Chlamydia trachomatis, smoking, HLA, HPV6/11) and non-causal (herpes simplex virus type 2) factors are put into perspective based on longitudinal, population-based studies. The established evidence base is now available for the evaluation of artificial intelligence/ machine learning performance in disclosing and judging causes of a chronic disease, cervical cancer. 

Seizing the day is pivotal in vaccination licensure studies, especially when the new vaccine is supposed to protect against a chronic infection with long lead time between the preventable infection and diagnosis of the to-be prevented chronic disease. With appropriate population-based design, that relied on the unique Finnish personal identification number comprehensive health register follow-up was feasible for the definition of safety, immunogenicity and efficacy of both the bivalent and quadrivalent human papillomavirus (HPV) vaccines soon after their licensure. In essentially HPV vaccination naïve population head-to-head comparison of the two vaccines was also feasible. Respectively, in 2002 and 2004 enrolled 1,749 and 4,809 adolescent girls around the ages of 16-17 years were respectively participated in two phase III licensure trials (FUTURE II and PATRICIA) of the quadrivalent and bivalent HPV vaccines. At the same time in 2003 and 2005, 15,615 adolescent, 18-19 year old girls from adjacent birth cohorts were enrolled into a concomitant control cohort. Linkage of the HPV vaccinees cohort with the population-based Finnish Maternity Cohort Serum Bank enabled comparative head-to-head studies on the quadrivalent and bivalent vaccineinduced total and neutralizing antibody responses, which were proven to be equally sustainable up to 12 years post-vaccination, however, with a logarithmic difference in the antibody levels. Linkage of the HPV vaccinees cohort and the concomitant control cohorts with the country-wide Finnish Cancer Registry has enabled the definition of vaccine efficacy (VE) against invasive cervical cancer and cervical intraepithelial neoplasia grade 3 (CIN3+) during 18 years of follow-up with comparable intention-totreat VEs of 68.4% and 64.5%. Linkage with the Hostital Discharge Registry has provided a sentinel, most notably for new onset autoimmune diseases (NOADs) that proved to be more than twice as sensitive as reporting of serious adverse effects but as such did not identify any NOAD-incidence differences between the HPV and control vaccines or unvaccinated population.

The overall impact of vaccinating a population, i.e., overall effectiveness always comprises direct effectiveness (vaccine efficacy) and indirect effectiveness (herd effect) of vaccination. In the case of human papillomavirus (HPV) vaccination the role of herd effect is especially strong due to the assortative nature of sexual risktaking behaviour and transmission of sexually transmitted micro-organisms, including genital HPV types. At the time of HPV vaccine licensure we launched a communityrandomized trial in Finland to provide real-life evidence on the impact of different vaccination strategies one of which was to be implemented into national Finnish vaccination program. In collaboration with the Finnish National Institute for Health and Welfare we invited in Autumn 2007 all 80,272 boys and girls from 1992-1995 birth cohorts, who attended one of the 250 junior high-schools in overall 33 Finnish towns outside the Helsinki Metropolitan regions. The study sites represented 33 of 34 Finnish communities with >35,000 inhabitants and were randomized 1:1:1 into 11 genderneutral HPV-vaccination communities, 11 girls-only HPV-vaccination communities, and 11 control (hepatitis B-virus) vaccination communities. Furthermore, with both parental and their own informed consent 20,513 girls and 11,662 boys participated in the school years 2007-2009. Between 2010 and 2014 11,396 cervical samples for HPV typing were obtained from 18.5 year-old females. We identified superb herd effect in the gender-neutral HPV vaccination communities against transient HPV18/31/33/35 infection as defined by PCR positivity and against persistent HPV type 16/18/31/35 positivity as defined by serology. Statistically significant rapid elimination of HPV types 18/31/33 by birth cohort was found only in the gender-neutral HPV vaccination communities. This study was possible only in the HPV vaccination naïve population, and the findings supported the implementation of gender-neutral HPV vaccination two years after their publication.

We anticipated that moderate human papillomavirus (HPV) vaccination in the Finnish community-randomized trial will demonstrate changes in ecological niche of the circulating HPVs such as the niche occupation by non-vaccine-covered HPV types. In this study we exploited a re-randomized cervical screening trial that had been launched in 2014 among 14,686 HPV-vaccinees starting from 22-year-old omen, born in 1992 (refer to Chapter V). Approximately half of the HPV-vaccinees, 6,958 women participated the trial and provided serial cervical samples for HPV typing at ages 22, 25 and 28 years. With the trial follow-up we could verify the more (gender-neutral vaccination arm) or less (girls-only vaccination arm) efficient vacation of ecological niche initially occupied by the vaccine-covered HPV types 16/18/31/45 up to eight years post-vaccination. However, no consistent changes were observed neither in the gender-neutral arm nor in the girls-only arm in epidemiological analyses for the nonvaccine-covered HPV39/51/52/56/58/59/66/68/73 determined by PCR of cervical samples or by serology. In contrast, our analyses at the community level revealed rising ecological diversity of the more or less non-vaccine-covered HPV types 33/35/51/52/56/58/59 in gender-neutral vaccination communities with a stronger herd immunity compared with girls-only vaccination communities from four to eight years post vaccination. This is probably the first recorded sign of niche occupation by the non-vaccine-targeted HPV types post-population level vaccination.

 We have been running an individually randomized cervical screening trial since 2014. All 14,686 HPV-vaccinated women from birth cohorts 1992-1995, who had received three vaccine shots between 12 to 15 years of age (12,402) or at 18 years of age (2,284). They were invited to participate in an individually randomized trial on infrequent vs.frequent cervical screening visits at ages 22, 25 and 28. The infrequently screened arm participants are informed only on cytological findings indicative of colposcopy and conisation, i.e., high- grade squamous cervical intraepithelial lesion (HSIL) or adenocarcinoma in situ (AIS). Furthermore, due to in-country migration female residents in one of the original 33 vaccination trial communities or after having moved to the Helsinki Metropolitan Region after 2014 were eligible. Altogether 6,958 women consented with a very high (over 92%) compliance to participate in the second (6,381 women) and the third (4,616 of 5,100 women vaccinated as early adolescents in 2007-2009) screening visits. The occurrence of cervical lesions: ASCUS, low-grade squamous intraepithelial lesion (LSIL), high-grade squamous intraepithelial lesion (HSIL) and adenocarcinoma in situ has been equal in the different arms. The progression potential of the HSIL findings in the HPV16/18 vaccinated women is probably reduced as suggested by the identification of hypermethylation of HPVindependent cervical cancer risk genes in only a few of the diagnosed HSIL cases. A randomized trial to compare mere clinical follow-up vs.treatment of HSILs diagnosed in vaccinated women is highly warranted.

During the last 30 years the incidence of head and neck cancers associated with human papillomavirus type 16 (HPV16) has rapidly increased in Finland and is now three-fold higher compared to 1990’s also in neighbouring Scandinavian countries, most notably Sweden. This is due to the widespread increase of HPV16 infection that started in the 1980’s both in Finland and Sweden. While, the recently launched, school-based gender-neutral HPV vaccination will eventually eliminate HPV-associated head and neck cancers preventing up to 1,000 annual cases in the Nordic countries, there are at least 30 birth cohorts that urgently need screening and treatment of the HPV-associated head and neck cancers. This chapter first evaluates the prospects of the elimination of HPV-associated head and neck cancers by genderneutral HPV-vaccination. Thereafter a proof-of-concept screening study that exploits the population-based FMC Serum Bank comprising first-trimester blood samples which were collected for serological screening from virtually all, 96% of the one million Finnish women, who were pregnant between 1983 and 2016 is characterized.