Analysis and Modeling to evaluate influenza vaccine responses

Slides: https://www.andreashandel.com/presentations/

Andreas Handel

2023-06-15

Acknowledgements

Zane Billings - for doing most of the work
Ted Ross, Andrea Sant - for data sharing
Many other colleagues and collaborators
NIH

Influenza heterogeneity

Factors that influence influenza vaccine responses

How do we define a vaccine response?

Pre-vaccination titer and vaccine response

Introduction

More pre-vaccination titer leads to more post-vaccination titer, but reduced boost.
Called response blunting, antibody ceiling effect, negative interference.
For a nice summary, see Oidtman et al. 2021 Trends Mic

Schematic

Pre-vaccination titer and vaccine response

Left: H1N1 vaccine, Right: H5N1 vaccine. Zarnitsyna et al 2015 Phil Trans B.

Pre-vaccination titer and vaccine response

Pre-vaccination titer and vaccine response

x-axis: titer increase. Left: H1N1, Right: H3N2. Moritzky et al 2023 JID.

UoR cohort - IgG titer increase

UoR cohort - HAI titer increase

H1N1 Pre-vaccination titer and boost

H3N2 Pre-vaccination titer and boost

Epitope-masking can explain some of those patterns

Zarnitsyna et al 2015 Phil Trans B, 2016 Plos Path.

Modeling pre-vaccination titer and heterologous vaccine response

Extension of the model can be used to explore predictions for heterologous response.

\[ \begin{aligned} \textrm{free antigen (type 1)} \qquad \dot{H_f} & = - k_1 A_1 H_f - k_2 A_2 H_f - d_f H_f \\ \textrm{bound antigen (type 1)} \qquad \dot{H_b} & = k_1 A_1 H_f + k_2 A_2 H_f - d_b H_b \\ \textrm{B-cell type 1} \qquad \dot{B_1} & = \frac{s_1 B_1 H_f}{p_1 + H_f} \\ \textrm{antibody type 1} \qquad \dot{A_1} & = g_1 B_1 -k_1 A_1 H_f - d_1 A_1\\ \textrm{B-cell type 2} \qquad \dot{B_2} & = \frac{s_2 B_2 H_f}{p_2 + H_f} \\ \textrm{antibody type 2} \qquad \dot{A_2} & = g_2 B_2 -k_2 A_2 H_f - d_2 A_2 \\ \end{aligned} \]

With \(k_2 = k_1(1-AD)\), \(s_2 = s_1(1-AD)\)

Modeling pre-vaccination titer and heterologous vaccine response

Extension of the model can be used to explore predictions for heterologous response for different antigenic distance (AD).

Heterologous vaccine response patterns - H1N1

Heterologous vaccine response patterns - H1N1

Heterologous vaccine response patterns - H1N1

Heterologous vaccine response patterns - H3N2

Heterologous vaccine response patterns - H3N2

Heterologous vaccine response patterns - H3N2

Epitope-masking can’t be the whole story

Pattern also seen for CD4 T-cells.

x-axis: titer increase. Left: H1N1, Right: H3N2. Moritzky et al 2023 JID.

Next steps

More careful comparison of model and data (fitting).
Extension of existing models to account for observed patterns in data.
Think about other important features (e.g., OAS/imprinting) and try to account for them in models.

Conceptual Framework

Bonus Slides

The role of prior vaccinations

There is some evidence that repeat vaccinations leads to reduced vaccine efficacy (VE).
Is the reduced VE mediated by antibody titers?
Repeat vaccinators mount a less strong vaccine response. But is that just because they start at a higher pre-vaccination titers?