Sustainability of neutralising antibodies induced by bivalent or quadrivalent HPV vaccines and correlation with efficacy: a combined follow-up analysis of data from two randomised, double-blind, multicentre, phase 3 trials

Background Quadrivalent and bivalent vaccines against oncogenic human papillomavirus (HPV) are used worldwide with different reported overall efficacies against HPV infections. Although protective concentrations of vaccine-induced antibodies are still not formally defined, we evaluated the sustainability of neutralising antibodies in vaccine trial participants 2–12 years after vaccination and the correlation with reported vaccine efficacy. Methods We did a follow-up analysis of data from the Finnish cohorts of two international, randomised, double-blind, phase 3 trials of HPV vaccines, PATRICIA (bivalent, HPV16 and 18) and FUTURE II (quadrivalent, HPV6, 11, 16, and 18). In 2002 and 2004–05, respectively, Finnish girls aged 16–17 years participated in one of these two trials and consented to health registry follow-up with the Finnish Cancer Registry. The cohorts were also linked with the Finnish Maternity Cohort (FMC) that collects first-trimester serum samples from nearly all pregnant Finnish women, resulting in 2046 post-vaccination serum samples obtained during up to 12 years of follow-up. We obtained serum samples from the FMC-based follow-up of the FUTURE II trial (from the quadrivalent vaccine recipients) and the PATRICIA trial (from corresponding bivalent vaccine recipients who were aligned by follow-up time, and matched by the number of pregnancies). We assessed neutralising antibody concentrations (type-specific seroprevalence) to HPV6, 16, and 18, and cross-neutralising antibody responses to non-vaccine HPV types 31, 33, 45, 52, and 58 from 2 to 12 years after vaccination. Findings Up to Dec 31, 2016, we obtained and analysed 577 serum samples from the quadrivalent vaccine recipients and 568 from the bivalent vaccine recipients. In 681 first-pregnancy serum samples, neutralising antibodies to HPV6, 16, and 18 were generally found up to 12 years after vaccination. However, 51 (15%) of 339 quadrivalent vaccine recipients had no detectable HPV18 neutralising antibodies 2–12 years after vaccination, whereas all 342 corresponding bivalent vaccine recipients had HPV18 neutralising antibodies.. In seropositive quadrivalent vaccine recipients, HPV16 geometric mean titres (GMT) halved by years 5–7 (GMT 3679, 95% CI 2377 to 4708) compared with years 2–4 (6642, 2371 to 13 717). Between 5 and 12 years after vaccination, GMT of neutralising antibodies to HPV16 and 18 were 5·7 times and 12·4 times higher, respectively, in seropositive bivalent vaccine recipients than in the quadrivalent vaccine recipients. Cross-neutralising antibodies to HPV31, 33, 45, 52, and 58 were more prevalent in the bivalent vaccine recipients but, when measurable, sustainable up to 12 years after vaccination with similar GMTs in both vaccine cohorts. Seroprevalence for HPV16, 31, 33, 52, and 58 significantly correlated with vaccine efficacy against persistent HPV infections in the bivalent vaccine recipients only (rs=0·90, 95% CI 0·09 to 0·99, p=0·037, compared with rs=0·62, 95% CI –0·58 to 0·97, p=0·27 for the quadrivalent vaccine recipients). Correlation of protection with prevalence of neutralising or cross-neutralising HPV antibodies was not significant in the quadrivalent vaccine recipients. Interpretation The observed significant differences in the immunogenicity of the two vaccines are in line with the differences in their cross-protective efficacy. Protective HPV vaccine-induced antibody titres can be detected up to 12 years after vaccination. Funding Academy of Finland and Finnish Cancer Foundation.