Functional analyses of vitt patient-derived PF4-antibodies and adenoviral vaccine compounds reveal unique patient-specific profiles of multicellular activation Free

Vaccine-induced thrombotic thrombocytopenia (VITT) is a rare but life-threatening adverse reaction to COVID-19 vaccines resulting in thrombosis and thrombocytopenia; particularly due to adenoviral vector-based vaccines (ChAdOx1 nCoV-19, Ad26.COV2.S). VITT is caused by antibodies against platelet factor 4 (anti-PF4) which activate platelets via FcγRIIa. Neutrophils and endothelial cells have also been suggested to be relevant in VITT. Clinical heterogeneity between VITT patients is observed but not understood from a functional and pathophysiologic point of view. We hypothesized that functionally different subsets of VITT anti-PF4 may be responsible for this heterogeneity, with varying impact on multiple cell types.

In this study, we aimed to characterize the immune activation effects of various VITT anti-PF4 as well as of adenoviral vaccine compounds, on platelets, neutrophils and endothelial cells (HUVECs). We utilized an unique set of 8 different recombinant reverse-engineered PF4-antibodies derived from serum of 4 VITT patients: patient 1: CR22044, CR22090, CR22091; patient 2: CR22045, CR22046; patient 3: CR22047; patient 4: CR22051, CR22066. (Wang et al, Blood 2024).

We first assessed both the dose kinetics of the antibodies (range of 10-500 ug/ml) as well as PF4 (10 and 100 ug/ml) with respect to platelet activation using flow cytometry (CD62p; P-selectin and Annexin V). We found that both the antibody and PF4 concentration determined the effect of platelet activation, to a varying extent depending on the specific antibody. Interestingly, the addition of 100 ug/ml PF4 could induce potent platelet activation occurring at lower antibody concentrations for some antibodies (e.g. CR22066, CR22090), while only occurring at higher concentrations for other antibodies (e.g. CR22045). Furthermore, we also found that some antibodies could significantly trigger FcγRIIa-dependent platelet-PF4 secretion (e.g. CR22046, CR22051), while other antibodies (e.g. CR22044, CR22090) could not. We found this to be in line with the functional FcγRIIa-dependent PF4-induced platelet activation assay (PIPAA) which assesses platelet aggregation.

Next, we assessed VITT-antibody induced neutrophil activation using flow cytometry (CD62L, CD66b, CD11b, MPO). Similarly to platelets, neutrophil activation was also dependent on dose kinetics of the antibodies and PF4. Neutrophil activation was observed, but not for all antibodies (CR22047). In addition, the ability of VITT antibodies to trigger platelet activation vs neutrophil activation was also not always in agreement, as antibody CR22044 could not activate platelets but did trigger neutrophil activation. Our preliminary data indicates that antibody-mediated platelet activation does not always translate to neutrophil extracellular traps, as measured by live-cell imaging.Endothelial cells were also stimulated with VITT antibodies, with and without pre-priming with LPS or TNF-alpha, and analyzed for activation markers (ICAM, VCAM, VE-Cadherin, E-selectin and P-selectin). Preliminary data do not support endothelial cell activation in our hands.

Platelets, neutrophils and endothelial cells were also exposed to adenoviral vaccine vector Ad26.Cov2.S. While an activating contribution was observed on all cell types, this appeared to be related to the clinical buffer (AP1 buffer). We subsequently analyzed the components of AP1 buffer, and could demonstrate that the observed activation was due to HBCD (a stabilizer), which we confirmed by comparing the effect of AP1 buffer vs AP1 buffer lacking HBCD vs HBCD alone. The effect of the adenoviral vaccine vector with VITT antibodies is currently being analyzed.

We also evaluated if spike protein could directly activate platelets. Upon exposure to spike protein, we did not observe any increase in platelet CD62p, nor any platelet-PF4 secretion, nor platelet aggregation in the functional PIPAA. Furthermore, we did also not observe any spike-protein induced cellular activation of neutrophils or endothelial cells. We will also test this in conjunction with VITT antibodies.

Taken together, we have constructed a blueprint of VITT anti-PF4-induced multicellular activation which reveals unique VITT patient-specific features, as demonstrated using 8 VITT-antibodies derived from 4 different patients. This advances our understanding of VITT antibody heterogeneity from a functional perspective, which may be exploited for novel diagnostic and therapeutic strategies.