Comprehensive Data Integration Approach to Assess Immune Responses and Correlates of RTS,S/AS01-Mediated Protection From Malaria Infection in Controlled Human Malaria Infection Trials

William Chad Young, Lindsay N. Carpp, Sidhartha Chaudhury, Jason A. Regules, Elke S. Bergmann-Leitner, Christian Ockenhouse, Ulrike Wille-Reece, Allan C. deCamp, Ellis Hughes, Celia Mahoney, Suresh Pallikkuth, Savita Pahwa, S. Moses Dennison, Sarah V. Mudrak, S. Munir Alam, Kelly E. Seaton, Rachel L. Spreng, Jon Fallon, Ashlin Michell, Fernando Ulloa-MontoyaMargherita Coccia, Erik Jongert, Galit Alter, Georgia D. Tomaras, Raphael Gottardo

Research output: Contribution to journalArticlepeer-review

Abstract

RTS,S/AS01 (GSK) is the world’s first malaria vaccine. However, despite initial efficacy of almost 70% over the first 6 months of follow-up, efficacy waned over time. A deeper understanding of the immune features that contribute to RTS,S/AS01-mediated protection could be beneficial for further vaccine development. In two recent controlled human malaria infection (CHMI) trials of the RTS,S/AS01 vaccine in malaria-naïve adults, MAL068 and MAL071, vaccine efficacy against patent parasitemia ranged from 44% to 87% across studies and arms (each study included a standard RTS,S/AS01 arm with three vaccine doses delivered in four-week-intervals, as well as an alternative arm with a modified version of this regimen). In each trial, RTS,S/AS01 immunogenicity was interrogated using a broad range of immunological assays, assessing cellular and humoral immune parameters as well as gene expression. Here, we used a predictive modeling framework to identify immune biomarkers measured at day-of-challenge that could predict sterile protection against malaria infection. Using cross-validation on MAL068 data (either the standard RTS,S/AS01 arm alone, or across both the standard RTS,S/AS01 arm and the alternative arm), top-performing univariate models identified variables related to Fc effector functions and titer of antibodies that bind to the central repeat region (NANP6) of CSP as the most predictive variables; all NANP6-related variables consistently associated with protection. In cross-study prediction analyses of MAL071 outcomes (the standard RTS,S/AS01 arm), top-performing univariate models again identified variables related to Fc effector functions of NANP6-targeting antibodies as highly predictive. We found little benefit–with this dataset–in terms of improved prediction accuracy in bivariate models vs. univariate models. These findings await validation in children living in malaria-endemic regions, and in vaccinees administered a fourth RTS,S/AS01 dose. Our findings support a “quality as well as quantity” hypothesis for RTS,S/AS01-elicited antibodies against NANP6, implying that malaria vaccine clinical trials should assess both titer and Fc effector functions of anti-NANP6 antibodies.

Original languageEnglish (US)
Article number672460
JournalFrontiers in Big Data
Volume4
DOIs
StatePublished - Jun 15 2021

Keywords

  • correlates of protection
  • immune response
  • machine learning
  • malaria
  • vaccine

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science (miscellaneous)
  • Information Systems

Fingerprint

Dive into the research topics of 'Comprehensive Data Integration Approach to Assess Immune Responses and Correlates of RTS,S/AS01-Mediated Protection From Malaria Infection in Controlled Human Malaria Infection Trials'. Together they form a unique fingerprint.

Cite this