Vaccine effectiveness against HPV infections

Abstract

Infections with human papilloma viruses (HPV) are the most prevalent sexually transmitted diseases. The vast majority of the sexually active population acquires them during lifetime but almost all HPV infections become resolved within a year. A small fraction of infections persists and can lead to cervical lesions, which may progress to cancer and their appearance is controlled by long-established cervical screening programs. Since licensure of the first HPV vaccines almost 15 years ago, more than one hundred countries introduced HPV vaccination into their national immunization strategies in order to prevent cervical cancer development. Despite permanent evaluation of immunological, virological and clinical endpoints of vaccine effectiveness, the antibody dynamics and protection are not fully understood, which is particularly true after natural infection. We thus studied humoral immune responses after natural infection and vaccination. In addition, we investigated reasons for rare events of cervical lesion development after immunization. In studies of serological responses to HPV infections (papers I and II) we demonstrated that the characteristics of type-specific anti-HPV antibodies to the most common oncogenic types (HPV 16 and 18) apply also to the other high-risk HPVs. Serum antibodies were generally stable over time and were strongly associated with cervical HPV DNA positivity. Seropositivity to multiple anogenital HPV types was also associated with presence of HPV DNA and we observed an association with an abnormal cytology and with the presence of non-genital HPVs. The latter finding suggests that a definite subset of women is more likely to seroconvert to multiple HPV types, which could be true also in induction of antibody responses to HPV vaccines.

Our serology assay was validated using HPV DNA from serially collected cervical samples as a standard. This was shown to be a valuable approach for evaluation of assay performance since repeated sampling for transient exposure may improve sensitivity without impairing specificity. In consequence, our multiplexed HPV serology assay was expected to be useful in studies of immune responses to HPV infection and vaccines. Our research on immunogenicity of the HPV vaccines was a head-to-head comparison of the anti-HPV antibody responses between the quadrivalent Gardasil and bivalent Cervarix (papers III and IV). We evaluated the sustainability of vaccine-targeted and cross-reactive serological responses up to 12 years post-vaccination. We found that the antibody levels induced by Cervarix were stable and virtually all women had anti-HPV antibody titers above those induced by HPV infection. In contrast, most of the recipients of quadrivalent Gardasil had detectable antibody levels but 8 % (HPV-16) and 18 % (HPV-18) showed a titer-decline to below the natural infection antibody level. Comparison of the antibody responses exhibited that the anti-HPV-16 antibody levels were 5.1-fold higher and anti-HPV-18 were 18.4-fold higher in the recipients of bivalent Cervarix (p < 0.0001). Further, the seropositivity to most of the non-vaccine HPV types was more prevalent among the women vaccinated with Cervarix than with the Gardasil. Nonetheless, anti-HPV antibody levels and avidity were comparable for almost all HPV types. For both vaccines, we found that the seroprevalence to cross-reactive types (HPV-31, -33, -35, -45, -52, -58, -59, -68 and -73) increased with the anti-HPV-16 levels. In the recipients of the bivalent HPV vaccine, antibody levels of the non-vaccine types HPV-31, -35, -51, -52, -56 and -58 increased with the anti-HPV-16 levels, which was not observed in women vaccinated with the Gardasil. Furthermore, our investigations showed that anti-HPV antibody levels shortly after vaccination predicted antibody levels and antibody avidity a decade later.

Lastly, we investigated whether the development of cervical lesions after HPV vaccination is associated with the vaccine-targeted HPV types (paper V). In this research project, we studied a cohort of women immunized with HPV vaccine before 17 years of age. We observed that not a single woman was positive for the low-risk vaccine types HPV-6 and -11. We found that cervical lesions were usually not associated with the vaccine-targeted HPV types of the highest oncogenic risk. Instead, non-vaccine types having a lower potential for inducing progression to cervical cancer were more likely to be present. Evidently, the latter finding has significant implications for the future design of cervical screening strategies that would target highly vaccinated cohorts of women and for design of surveillance and monitoring strategies. Another consequence of these findings is the information to physicians and vaccinated women that cervical dysplasia may occur despite HPV vaccination but likely caused by HPV types that rarely progress to cancer.

In summary, the findings presented in this thesis contribute to the understanding of the dynamics and the durability of naturally acquired and vaccine-induced antibody responses to HPV. The results give reasons for the cervical lesion-development among young HPV-vaccinated women. We conclude that the future definition of a clinically relevant anti-HPV antibody protective threshold, as well as the evaluation of determinants for higher and broader immunological responses to HPV vaccines are necessary to optimize HPV vaccine.