Estimation of primary, secondary and composite effects of the RTSS/AS01 vaccine

This work is a joint research project between Professor Yin Bun Cheung, Duke-NUS Medical School, Singapore; Professor Eddy EK Lam, University of Hong Kong, and Professor Paul Milligan at the London School of Hygiene and Tropical Medicine. We propose to develop and evaluate statistical methods and tools for estimating primary secondary and composite effects of malaria vaccines, using the phase 3 trial data as an example. We will estimate vaccine efficacy using all clinical malaria episodes, using the same case definitions as in the original trial protocol. For estimation of the total effect - i.e. incorporating the secondary effect via influence on accumulation of partial immunity - we will use the Andersen-Gill model [1]. To estimate the primary effect of the vaccine - i.e. controlling for the effect of accumulation of partial immunity - we will stratify for number of previous clinical malaria episodes. Since the stratification induces a selection bias, we will include a frailty factor to control for the selection bias [2,3]. We will also investigate methods to allow for the duration of immunity (such that earlier events have less influence than more recent events) when estimating these effects. We will use the phase 3 trial data as the motivating example, using these data to develop and evaluate the methods, and to demonstrate the use of the methods. The estimation of vaccine efficacy often only uses person-time and event outcomes 14 days after the injection of the last dose of the vaccine. Data between the date of the first dose to 14 days after the last dose is usually not used in the analysis, because the vaccine is not supposed to generate sufficient immune response during this period. We propose an analytic approach to answer the questions of event dependency and primary effect by a joint modelling approach with an AR(k) structure and fully utilizing the data from the date of first dose. To be concrete, we will use the data of the phase 2 trial of the RTS,S/AS02 malaria vaccine to illustrate. In an additional analysis, we will use the phase 3 and phase2 data to determine if there is any evidence that malaria incidence differs by gender, and whether the total and primary effects of the RTS,S/AS01 or RTS,S/AS02 vaccines differ by gender, using the methods described above. We will also investigate whether any gender difference is influenced by age at vaccination, and by receipt of other vaccines received prior to vaccination with RTS,S/AS01.1. Cheung YB, Xu Ying, Tan Sze Huey, Cutts F, Milligan PJM (2010) Estimation of intervention effects using first or multiple episodes in clinical trials: The Andersen-Gill model re-examined. Statistics in Medicine 29(3):328-336.2. Xu Y, Cheung YB, Lam E, Milligan P (2012) Estimation of summary protective efficacy using a frailty mixture model for recurrent event time data. Statistics in Medicine, 31(29):4023-39. 3. Cheung YB, Ma X, Lam KF, Milligan P (2020) Estimation of the primary, secondary and composite effects of malaria vaccines using data on multiple all clinical malaria episodes.