Analyses of health outcomes from the 5 sites participating in the Africa and Asia clinical efficacy trials of the oral pentavalent rotavirus vaccine

Abstract

Efficacy of the pentavalent rotavirus vaccine (PRV), RotaTeq(®), against severe rotavirus gastroenteritis (RVGE) was evaluated in two double-blind, placebo-controlled, multicenter Phase III clinical trials conducted in GAVI-eligible countries in Africa (Ghana, Kenya, and Mali) and in Asia (Bangladesh and Vietnam) from March 2007 through March 2009. The findings from each continent have been analyzed and presented separately, according to a single identical protocol. Ad hoc analyses combining data from the five sites were performed to further assess the impact of PRV. 6674 infants (4705 infants from Africa and 1969 infants from Asia), randomized 1:1 to receive 3 doses of PRV/placebo at approximately 6-, 10-, and 14-weeks of age according to each country's EPI schedule, were included in the per protocol efficacy analysis. Breastfeeding and concomitant administration of EPI vaccines, including OPV, were allowed. Episodes of gastroenteritis (GE) in infants who presented to study facilities were captured and scored using the 20-point Vesikari scale. Stool samples were analyzed by rotavirus-specific EIA to detect presence of rotavirus antigen and RT-PCR to determine the G/P genotypes. We assessed efficacy to prevent all-cause GE and RVGE at a variety of cut-off points (score≥11, severe; score≥15, very severe). Vaccine efficacy (VE) against RVGE, regardless of serotype, through the entire follow-up period for any severity, severe (score≥11), and very severe (score≥15) was 33.9%, 95% CI (22.7, 43.5), 42.5%, 95% CI (27.4, 54.6), and 51.2%, 95% CI (26.3, 68.2), respectively. Through the first year of life, VE against severe RVGE was 58.9%, 95% CI (40.0, 72.3) and against all-cause severe GE was 23.0%, 95% CI (5.4, 37.3). VE against severe RVGE caused by non-vaccine G serotypes, G8 and G9, through the entire follow-up period was 87.5%, 95% CI (6.8, 99.7) and 48.0%, 95% CI (-5.5, 75.6), respectively. All G8 strains were associated with P2A[6] (a P-type not contained in PRV), while the majority of the G9 strains were associated with P1A[8] (a P-type contained in PRV). Combining data from the 5 sites strengthens the precision of VE estimates and reveals rising VE with increased RVGE severity. Extrapolating data from VE against severe GE and RVGE suggest that 39% of severe GE episodes during the first year of life were due to rotavirus, highlighting substantial, potentially preventable, public health burden of RVGE. PRV provides protection against non-vaccine serotypes (G8P2A[6]).