The Journal of Immunology
May 1, 2016 vol.196

.
A multivariate approach to data analysis of vaccine clinical trials.
M Coccia1, F Nozay1, L De Mot1, Avisek Deyati1, E Jogert1, R van der Most1 and R van den Berg1
Author Affiliations
1GlaxoSmithKline, Belgium
Abstract
Despite significant progress in prevention, diagnosis and treatment, Malaria and Tuberculosis (TB) remain major health challenges. In 2014 WHO estimated that ≈438000 people died from Malaria and ≈1.5 million from TB, mainly in resource-poor countries. Vaccines represent a cost-effective and efficient method of preventing infectious diseases. The development of vaccines for TB and Malaria would significantly contribute to reducing disease burden, particularly with the emergence of drug-resistant pathogens. GSK’s Malaria vaccine, Mosquirix™ (RTS,S/AS01), received a positive opinion from European regulators for the prevention of Malaria in young children in sub-Saharan Africa. GSK’s candidate vaccine for TB (M72/AS01) induces robust TB-specific CD4 T-cell responses in humans, and it is undergoing phase IIb clinical trials. System biology approaches can support vaccines at different stages of development by identifying molecular signatures that drive responses to vaccination. Here, we describe the application of systems vaccinology to the analysis of gene expression data from clinical trials for Malaria and TB candidate vaccines. Our analysis aimed to identify early predictors of vaccine efficacy in Malaria clinical trials and to detect signatures associated with reactogenicity in the early development of a TB vaccine. To better capture the multidimensional nature of the data, we used multivariate analysis approaches such as Partial Least Squares regression. Additionally, biological interpretation of our results allowed us to pinpoint biological processes linked with response to vaccination, advancing our understanding of vaccine mode of action.