High-dose influenza vaccine favors acute plasmablast responses rather than long-term cellular responses

Abstract

High-dose (HD) influenza vaccine shows improved relative efficacy against influenza disease compared to standard-dose (SD) vaccine in individuals ⩾65years. This has been partially credited to superior serological responses, but a comprehensive understanding of cell-mediated immunity (CMI) of HD vaccine remains lacking. In the current study, a total of 105 participants were randomly administered HD or SD vaccine and were evaluated for serological responses. Subsets of the group (n=12-26 per group) were evaluated for B and T cell responses at days 0, 7, 14 and 28 post-vaccination by flow cytometry or ELISPOT assay. HD vaccine elicited significantly higher hemagglutination inhibition (HI) titers than SD vaccine at d28, but comparable titers at d365 post-vaccination. HD vaccine also elicited higher vaccine-specific plasmablast responses at d7 post-vaccination than SD vaccine. However, long-lived memory B cell induction, cytokine-secreting T cell responses and persistence of serological memory were comparable regardless of vaccine dose. More strategies other than increased Ag amount may be needed to improve CMI in older adults.

Trial registration: ClinicalTrials.gov NCT 01189123.

Trial registration: ClinicalTrials.gov NCT01189123.

Keywords: Cell-mediated immunity; High-dose influenza vaccine; Older adults; Plasmablast response.

Conflict of interest statement

H Keipp Talbot received research support from Sanofi-Pasteur, Gilead and MedImmune. She currently serves as an advisor for Novartis and VaxInnate; Marie R Griffin received grant support from MedImmune and Pfizer (formerly Wyeth). All the remaining authors disclosed no potential conflicts of interest.

Published by Elsevier Ltd.

Figures

Figure 1. HD induced significantly higher serological responses than SD vaccine against all vaccine components, but did not persist up to a year

(A) GMT fold-rise (d28/d0 and d365/d0) was estimated to determine the vaccine-associated changes in Ab responses within HD vaccine recipients (d0, N=51; d28, N=49; d365, N=49) or SD vaccine recipients (d0, N=52; d28, N=51; d365, N=51) for all vaccine components. A GMT fold-rise of 1 (dotted line) is indicative of no vaccine-induced change. (B) The difference in responses between HD and SD group was estimated by HD/SD ratio at each time-point. A HD/SD ratio of 1 is indicative of no difference between HD and SD-induced Ab responses. (C) GMT fold-rises (d28/d0 and d365/d0) were estimated to determine the vaccine-mediated boost responses for the 2009–2010 influenza vaccine components (A/Bris59 (H1N1), A/Bris10 (H3N2) and B/Flor). (D) HD/SD ratio was estimated to determine the difference in boost responses between HD and SD group. Error bars indicate 95% confidence intervals. Statistical significance is shown as *p ≤ 0.05, **p ≤ 0.01, *** p ≤ 0.001.

Figure 2. HD induced significantly higher plasmablast responses at d7 post-vaccination

Plasmablast responses were analyzed by flow cytometry and ELISPOT assay using PBMCs (N=12) randomly selected from each vaccine group at d0 and d7 post-vaccination. (A) Plasmablasts were identified as CD20−CD3−CD27+CD38+ cells by flow cytometry and their percent in PBMCs were calculated to determine geometric mean percent (GMP) at each time-point. GMP fold-rise (d7/d0) was estimated to determine vaccine-induced responses. A GMP fold-rise of 1 (dotted line) is indicative of no vaccine-induced change. Estimated GMP values from each time-point were compared to determine the difference in response between HD vs. SD group. (B–D) Ab-secreting cells (ASCs) specific to A/Cal (H1), A/Perth (H3), or B (Bris) were measured by ELISPOT assay for IgG-secreting cells. GMP fold-rise (d7/d0) were estimated to determine vaccine-induced responses. Difference in responses between HD vs. SD group were determined by HD/SD ratio at each time-point. Error bars indicate 95% confidence intervals. Statistical significance is shown as *p ≤ 0.05, **p ≤ 0.01, *** p ≤ 0.001 or stated.

Figure 3. Cell-mediated immunity was comparable between HD and SD group

Paired sets (d0/d7/d14/d28) of PBMCs randomly selected from HD (n=26) and SD (n=22) were stimulated in vitro with MOI of 1 A/Cal (H1) or A/Perth (H3) wild-type virus or left unstimulated (unstim; u.s.) overnight and cytokine secretion from CD4 or CD8 T cells was analyzed by flow cytometry. GMP of activated (CD69+), IFNγ+(A) or TNFα+(B) CD4 T cells or IFNγ+(C) or TNFα+(D) CD8 T cells were estimated at days 0,7,14 and 28 post-vaccination. GMP fold-rises (d7/d0, d14/d0, d28/d0) were estimated to determine vaccine-induced changes and shown by virus subtype and vaccine dose. Comparison of HD and SD-induced CMI were estimated by HD/SD ratios at days 0, 7, d14 and d28. A GMP fold-rise of 1 (dotted line) is indicative of no vaccine-induced change, and a HD/SD ratio of 1 (dotted line) is indicative of no difference between HD and SD-induced T cell responses. Error bars indicate 95% confidence intervals. Statistical significance is shown by *p ≤ 0.05, **p ≤ 0.01, *** p ≤ 0.001.

Figure 4. Pre-existing CMI level correlated with post-vaccination CMI regardless of vaccine dose

The pre-existing CMI of HD and SD group was tested for correlations with their corresponding post-vaccination responses including CMI and Ab responses using a linear mixed model. (A) Pre-existing (d0) A/Cal (H1) or A/Perth (H3)-specific CD4+IFNγ+ or CD8+TNFα+ T cell responses were tested for correlations with d28 post-vaccine responses. (B) Post-vaccine (d28) A/Cal (H1) or A/Perth (H3)-specific CD4+IFNγ+ T cell responses were tested for correlations with d28 HI titers. r2 indicates correlation coefficient. Solid or dotted lines represent regression lines of HD or SD vaccine-induced responses, respectively.