Characterization of potential biomarkers of reactogenicity of licensed antiviral vaccines: randomized controlled clinical trials conducted by the BIOVACSAFE consortium

Abstract

Biomarkers predictive of inflammatory events post-vaccination could accelerate vaccine development. Within the BIOVACSAFE framework, we conducted three identically designed, placebo-controlled inpatient/outpatient clinical studies (NCT01765413/NCT01771354/NCT01771367). Six antiviral vaccination strategies were evaluated to generate training data-sets of pre-/post-vaccination vital signs, blood changes and whole-blood gene transcripts, and to identify putative biomarkers of early inflammation/reactogenicity that could guide the design of subsequent focused confirmatory studies. Healthy adults (N = 123; 20-21/group) received one immunization at Day (D)0. Alum-adjuvanted hepatitis B vaccine elicited vital signs and inflammatory (CRP/innate cells) responses that were similar between primed/naive vaccinees, and low-level gene responses. MF59-adjuvanted trivalent influenza vaccine (ATIV) induced distinct physiological (temperature/heart rate/reactogenicity) response-patterns not seen with non-adjuvanted TIV or with the other vaccines. ATIV also elicited robust early (D1) activation of IFN-related genes (associated with serum IP-10 levels) and innate-cell-related genes, and changes in monocyte/neutrophil/lymphocyte counts, while TIV elicited similar but lower responses. Due to viral replication kinetics, innate gene activation by live yellow-fever or varicella-zoster virus (YFV/VZV) vaccines was more suspended, with early IFN-associated responses in naïve YFV-vaccine recipients but not in primed VZV-vaccine recipients. Inflammatory responses (physiological/serum markers, innate-signaling transcripts) are therefore a function of the vaccine type/composition and presence/absence of immune memory. The data reported here have guided the design of confirmatory Phase IV trials using ATIV to provide tools to identify inflammatory or reactogenicity biomarkers.

Conflict of interest statement

P.D., R.v.d.B. and G.D.G. are employees of the GSK group of companies, and report receiving restricted shares of the company. K.K. is employee of Sanofi Pasteur. I.J. is employee of deCode genetics/Amgen Inc. A.M. and B.B. are receiving royalties for reagents to measure PTX3 levels. Other authors report no conflict of interest.

Figures

Figure 1

Oral temperature and pulse measurements after immunisation. The figure shows the group mean oral temperature (a), and pulse rate (b), recorded at various time points during the first 48 hours post vaccination for each treatment group. Ribbons indicate the SEM by group.

Figure 2

Changes in serum total protein and alanine transaminase concentrations post-immunization. The figure shows the mean concentrations of serum total protein (a; normal range 60–80 g/L) and alanine transaminase (b; normal range 0–49 U/L), measured on selected days after immunisation on Day 0. Ribbons indicate group SEM. Vertical red lines indicate the time of discharge from the residential unit.

Figure 3

Changes in serum CRP and PTX3 concentrations post-immunization. Mean concentrations of C-Reactive Protein (CRP) and pentraxin 3 (PTX3) were measured on selected days after immunisation on Day 0. Panel (a) Normal CRP range (0–10 mg/L) is indicated by horizontal dotted red lines. Inset panel: CRP concentrations in the ATIV, TIV and placebo groups were measured at frequent time points in the first 36 hours post immunisation. Panel (b) PTX3 (no normal range quoted) concentrations on Days 0–5, including frequent time points in the first 36 hours post immunisation are shown for the ATIV, TIV and placebo groups. Ribbons indicate group SEM. The vertical red line indicates the time of discharge from the residential unit.

Figure 4

Changes in serum IP-10 and MCP-1 concentrations post immunization. Levels of interferon γ-induced protein 10 (IP-10; panel a) and monocyte chemoattractant protein 1 (MCP1; panel b) were measured up to 28 days after immunization on Day (D)0. Values are expressed as group mean fold-change from baseline (D0) with SEM, to accommodate the different absolute concentration ranges. Vertical dotted lines indicate the time of discharge from the residential unit.

Figure 5

Blood transcriptomic responses for all groups (group level). Peripheral whole blood transcriptomics data are shown by vaccine group. Schematics represent the responses detected for modules of IFN signaling-associated genes, CD4 stimulation-related genes, and innate-cell (monocyte, neutrophil, dendritic cell [DC])-associated genes, as presented by blue, green and pink lines, respectively. Each line represents a single blood transcriptional module (BTM,). The Y-axis presents -log10(FDR) values. Vertical red lines indicate the time of discharge from the residential unit. The contrast tested for a given vaccine and a given time point was the interaction between the differences in expression between this time point and Day 0, and between the given vaccine and placebo.

Figure 6

Blood transcriptomic responses for the TIV and ATIV groups in Study C. Heatmap presentations of responses for the individual BTMs of subjects vaccinated with TIV and ATIV are shown. Numbers shown above the heatmap represent the time-points post vaccination, expressed in days.

Figure 7

Interindividual variability in transcriptomics profiles of the ATIV group. Legend as for Fig. 6. Each column corresponds to an individual subject (presented in the same sequence across the three days).

Figure 8

Correlation between serum IP-10 levels and IFN-related gene activation. (a) ATIV recipients were stratified into ‘early responders’ or ‘non-responders’ with respect to the presence or absence of early (Day [D]1) enrichment in the IFN signatures. Serum IP-10 levels categorized by responder status are presented for the ATIV and placebo groups. (b) Eigengene plot showing the correlation between the eigengene of the ‘Interferon’ blood transcriptional module (BTM) DC.M5.12 and the expression of IP-10 in the ATIV group at D0 and D1. Eigengene (as defined previously) of a set of genes is calculated based on the expression of all genes in a gene set (BTM) and is correlated with the expression of the majority of the genes in the BTM. Pearson’s r = 0.81; p < 1e-6.