Predictors of oral rotavirus vaccine immunogenicity in rural Zimbabwean infants

James A Church,Bernard Chasekwa,Sandra Rukobo,Margaret Govha,Benjamin Lee,Marya P Carmolli,Robert Ntozini,Kuda Mutasa,Monica M Mcneal,Florence D Majo,Naume V Tavengwa,Beth D Kirkpatrick,Lawrence H Moulton,Jean H Humphrey,Andrew J Prendergast

doi:10.1016/j.vaccine.2020.01.097

Abstract

BackgroundOral rotavirus vaccines (RVV) have poor immunogenicity in low-income countries, for reasons that remain unclear. This study identified the determinants of RVV immunogenicity among infants in rural Zimbabwe. MethodsAnti-rotavirus IgA titres were measured among a sub-group of infants enrolled in the Sanitation Hygiene Infant Nutrition Efficacy (SHINE) trial (NCT01824940). SHINE was a cluster-randomized trial of improved infant and young child feeding, and improved water, sanitation and hygiene (WASH) in two rural Zimbabwean districts. Infants received RVV as part of the national immunisation programme. Among HIV-unexposed infants in the non-WASH trial arms, we evaluated associations between potential risk factors (vaccine schedule and dose, maternal and infant nutritional status, infant diarrhoea, and household environment) and RVV immunogenicity (seroconversion, seropositivity and geometric mean titres) using multivariable regression. ResultsAmong 219 infants with seroconversion data, 43 (20%) successfully seroconverted and 176 (80%) failed to seroconvert to RVV. Seroconversion was positively associated with a higher length-for-age Z-score (LAZ) around the time of vaccination (adjusted RR 1.27 (95% CI 1.04, 1.55), P = 0.021), and negatively associated with concurrent OPV and RVV administration (adjusted RR 0.36 (0.19, 0.71), P = 0.003). Among 472 infants with post-vaccination titres, a higher LAZ score was associated with increased seropositivity (aRR 1.21 (95% CI 1.06, 1.38), P = 0.004), and higher birthweight was associated with increased IgA titres (0.45 (95%CI 0.18, 1.09) U/mL greater per 100 g gain in birthweight; P = 0.001). ConclusionsInfant ponderal and linear growth were positively associated with RVV immunogenicity, while concurrent administration of OPV was negatively associated with RVV immunogenicity. Together, these findings suggest that improving foetal growth and separating RVV and OPV administration are plausible approaches to increasing RVV immunogenicity.

Highlights

Oral rotavirus vaccines (RVV) have poor immunogenicity in low-income countries, for reasons that remain unclear
We have previously shown that infants randomised to receive the WASH intervention had improved RVV immunogenicity [15], so infants in the WASH arms were excluded from this analysis to allow for unbiased evaluation of the predictors of RVV immunogenicity
472 infants fulfilled the inclusion criteria for this analysis: all were in non-WASH arms with a complete record of RVV receipt and RVV immunogenicity data available

Summary

Introduction

Oral rotavirus vaccines (RVV) have poor immunogenicity in low-income countries, for reasons that remain unclear. Among HIV-unexposed infants in the non-WASH trial arms, we evaluated associations between potential risk factors (vaccine schedule and dose, maternal and infant nutritional status, infant diarrhoea, and household environment) and RVV immunogenicity (seroconversion, seropositivity and geometric mean titres) using multivariable regression. Seroconversion was positively associated with a higher length-for-age Zscore (LAZ) around the time of vaccination (adjusted RR 1.27 (95% CI 1.04, 1.55), P = 0.021), and negatively associated with concurrent OPV and RVV administration (adjusted RR 0.36 (0.19, 0.71), P = 0.003). Conclusions: Infant ponderal and linear growth were positively associated with RVV immunogenicity, while concurrent administration of OPV was negatively associated with RVV immunogenicity. Together, these findings suggest that improving foetal growth and separating RVV and OPV administration are plausible approaches to increasing RVV immunogenicity.

Objectives

Methods

Results

Conclusion