Design and Analysis of Vaccine Trials for Emerging Infectious Disease Threats

点击翻译按钮获取中文摘要

ABSTRACT Outbreaks of emerging pathogens have become more frequent, and we will likely face future epidemics for which we are not adequately prepared. Effective vaccines are a critical tool for controlling infectious disease threats, but experimental products must be rigorously evaluated for efficacy and safety before they can be licensed and deployed. Our experience with the 2013-2016 West African Ebola epidemic highlighted the unique challenges of conducting phase III vaccine efficacy trials for emerging pathogens. Besides issues of strained infrastructure in typically resource-limited settings, incidence may be highly heterogeneous in the population and spatiotemporally hard to predict. The outbreak may end before enough cases have accrued to establish efficacy, as occurred in two of the three phase III Ebola vaccine trials. Compared to standard vaccine clinical trials, researchers conducting trials during public health emergencies also generally have less information about the disease and/or vaccine. We supported the design and analysis of a third phase III Ebola vaccine trial in Guinea that used a novel ring vaccination approach. Modeled after the strategy used to eradicate smallpox, clusters were defined as the contacts and contacts of contacts of laboratory-confirmed Ebola virus disease cases and then cluster- randomized to immediate or delayed vaccination. This trial demonstrated high efficacy of the candidate vaccine, and its success was in part attributed to its innovative, responsive strategy that tracked the epidemic as it progressed, precisely targeting individuals at highest risk of exposure. These experiences have motivated an international call for novel methods for vaccine trial design and analysis adapted for emerging infectious disease threats. We propose the development of flexible, adaptive trial design strategies intended to increase the efficiency and likelihood of success when evaluating experimental vaccines in outbreak settings. Our first aim is to develop a class of responsive designs and associated tools for sample size calculation and analysis generalized beyond ring vaccination. Our second aim is to outline strategies for implementing trials in outbreaks of unpredictable duration, including how to define data monitoring rules and how to aggregate information across outbreaks when any given outbreak is too small to support a trial. Our third aim is to design adaptive, multi-arm vaccine trials with or without a control comparator. For each aim, we will describe the designs and the relevant qualitative considerations, develop and validate the supporting statistical methods, and evaluate their robustness using realistic, mathematical and computational disease transmission models. Our research represents a critical first step for evaluating these strategies before they could be implemented in the field.