Primary Submission Category: Mediation
Identifying correlate of protection for COVID-19 vaccines using causal inference framework
Authors: Nancy Hiu Lan Leung,
Presenting Author: Nancy Hiu Lan Leung*
The COVID-19 pandemic has resulted in significant loss of life, and vaccination has been instrumental in helping us return to a new normal. Immune biomarkers, for example neutralizing antibodies, are used to predict COVID-19 vaccine efficacy and to evaluate updated vaccines in immunobridging studies, i.e. these immune biomarkers are considered as correlate of protection (CoP). However, in a recent randomized trial of third-dose COVID-19 vaccination, we found that neutralizing antibodies did not associate with vaccine efficacy for inactivated COVID-19 vaccines. A significant hurdle to identify new CoP for inactivated vaccines is the difficulty to attribute causal effect of individual CoP to protection; but so far only limited research has been conducted to address this analytic challenge. We are conducting additional serologic testing of different immune biomarkers against circulating viruses in infected and uninfected individuals from our several on-going longitudinal cohorts and vaccine trials of COVID-19 vaccination in Hong Kong, and we also collect other relevant data such as demographics and prior vaccination and infection history. We will evaluate causal diagrams that reflect different hypotheses of immune protection with these data. We will evaluate the potential role of individual immune biomarkers as CoP by estimating the causal effect of vaccination on protection mediated by each biomarker.