Systems serology analysis shows IgG1 and IgG3 memory responses six years after one dose of quadrivalent HPV vaccine

Abstract

The WHO has given a permissive recommendation for an off-label one-dose human papillomavirus (HPV) vaccine schedule to prevent cervical cancer, based on evidence of comparable protection to two or three doses of vaccine. While neutralizing antibodies are thought to be the primary mechanism of protection, the persistence of immunity and whether other antibody-mediated mechanisms of protection are involved is unclear. Using systems serology, we investigated HPV antibody responses in serum from Fijian girls who were unvaccinated or received one, two or three doses of quadrivalent HPV vaccine six years earlier. We also evaluated their HPV antibody responses 28 days following a dose of bivalent HPV vaccine. After six years, one dose induced lower antibody concentrations but similar antibody profiles and phagocytic function as two or three doses. Following bivalent vaccine, antibody concentrations, particularly IgG1/IgG3, antibody profiles and phagocytic function were similar between previously vaccinated girls, indicating immune memory after one dose. Cross-reactive antibody responses against non-vaccine genotypes (HPV31/33/45/52/58) were lower following one dose than two or three doses. These findings provide novel insights into serological immunity and recall responses following one-dose HPV vaccination.

Fig. 1. HPV vaccine-type antibody subclass responses in Fijian girls.

Violin plots show total IgG, IgM, IgA1, IgA2 MFI for HPV16 (A) and HPV18 (C), and IgG subclass MFI i.e. IgG1, IgG2, IgG3, IgG4 for HPV16 (B) and HPV18 (D) median fluorescence intensity (MFI) or International Units/mL (IU/mL) six years following the last dose of 4vHPV vaccine (Pre) or 28 days post-2vHPV booster (Post). The median of each group is denoted by a black horizontal line. Statistical significance was calculated using the Kruskal-Wallis test and corrected for multiple comparisons with Dunn’s test. A positive antibody response is defined as being above median BPV (negative control antigen) MFI or IU/mL (IgG). Definitions: 0D 0-dose (n = 20), 1D 1-dose (n = 20), 2D 2-dose (n = 21), 3D 3-dose (n = 19).

Fig. 2. HPV vaccine-type Fc-mediated antibody responses in Fijian girls.

Fc receptor responses (FcγR2A, FcγR2B, FcγR3A) and ADCP function against HPV16 and HPV18 (A) were measured at six years following the last dose of 4vHPV vaccine (Pre) or 28 days post-2vHPV booster (Post). The median of each group is denoted by a black horizontal line. A positive response is defined as being above the median of control antigen BPV for Fc receptors, or the median of a seronegative control serum for ADCP. Statistical significance was calculated using the Kruskal-Wallis test and corrected for multiple comparisons with Dunn’s test. Correlation between FcγR2A and ADCP for HPV16 and HPV18 (B) was assessed using Spearman’s rank test. Definitions: 0D 0-dose (n = 20), 1D 1-dose (n = 20), 2D 2-dose (n = 21), 3D 3-dose (n = 19).

Fig. 3. HPV antibody signature in unvaccinated and vaccinated Fijian girls.

The polar plots depict the relative magnitude of HPV16 and HPV18 antibody features six years following the last dose of 4vHPV vaccine (Pre) or 28 days post-2vHPV booster (Post). Wedge sizes represent the median of each antibody feature, normalised within HPV genotypes and antibody features, and across all dose groups; 0-dose (n = 20), 1-dose (n = 20), 2-dose (n = 21), 3-dose (n = 19).

Fig. 4. Principal component analysis (PCA) of vaccine-type HPV antibody features across unvaccinated and 4vHPV vaccinated Fijian girls.

PCA analyses and loading plots of log-transformed, LASSO-selected HPV16/18 antibody features in unvaccinated (0D, n = 20) or 4vHPV-vaccinated individuals (1D, 1-dose, n = 20; 2D, 2-dose n = 21; 3D, 3-dose, n = 19) six years following their last dose of vaccine, and one month after a 2vHPV booster dose. The mean of each group is represented by the large solid circles in the PCA plot, with coloured ellipses representing the 95% confidence level of each dosage group. Bars on the loading plot show the contribution of each LASSO-selected antibody feature to PC1 in ascending order.

Fig. 5. Breadth of cross-reactive HPV antibody responses in unvaccinated and vaccinated Fijian girls.

Stacked bar plots represent the percentage of individuals in each of 0-dose (0D, n = 20), 1-dose (1D, n = 20), 2-dose (2D, n = 21) and 3-dose (3D, n = 19) 4vHPV vaccine dosage groups after six years (A), and following 2vHPV (B) with detectable responses to any of five non-vaccine genotypes (HPV31, HPV33, HPV45, HPV52, HPV58) for each antibody feature. Detectable antibody responses defined as having MFI > mean + 2 standard deviations of BPV MFI.