Genetic and phenotypic stability of poliovirus shed from infants who received novel type 2 or Sabin type 2 oral poliovirus vaccines in Panama: an analysis of two clinical trials

Abstract

Sabin strains used in oral poliovirus vaccines (OPV) can revert to virulence and, in rare instances, cause disease or generate vaccine-derived strains leading to outbreaks in areas of low immunisation coverage. A novel OPV2 (nOPV2) was designed to stabilise the viral genome against reversion and reduce recombination events that might lead to virulent strains. In this study, we evaluated the genetic and phenotypic stability of shed poliovirus following administration of one dose of monovalent OPV2 (mOPV2) or nOPV2 to infants aged 18-22 weeks. In two similarly designed clinical trials (NCT02521974 and NCT03554798) conducted in Panama, infants aged 18-22-weeks, after immunisation with three doses of bivalent OPV (types 1 and 3) and one dose of inactivated poliovirus vaccine, were administered one or two doses of mOPV2 or nOPV2. In this analysis of two clinical trials, faecally shed polioviruses following one dose of mOPV2 or nOPV2 were isolated from stools meeting predetermined criteria related to sample timing and viral presence and quantity and assessed for nucleotide polymorphisms using next-generation sequencing. A transgenic mouse neurovirulence test was adapted to assess the effect of the possible phenotypic reversion of shed mOPV2 and nOPV2 with a logistic regression model. Of the 91 eligible samples, 86 were able to be sequenced, with 72 evaluated in the transgenic mouse assay. Sabin-2 poliovirus reverts rapidly at nucleotide 481, the primary attenuation site in domain V of the 5' untranslated region of the genome. There was no evidence of neurovirulence-increasing polymorphisms in domain V of shed nOPV2. Reversion of shed Sabin-2 virus corresponded with unadjusted paralysis rates of 47·6% at the 4 log10 50% cell culture infectious dose (CCID50) and 76·7% at the 5 log10 CCID50 inoculum levels, with rates of 2·8% for 4 log10 CCID50 and 11·8% for 5 log10 CCID50 observed for shed nOPV2 samples. The estimated adjusted odds ratio at 4·5 log10 of 0·007 (95% CI 0·002-0·023; p<0·0001) indicates significantly reduced odds of mouse paralysis from virus obtained from nOPV2 recipients compared with mOPV2 recipients. The data indicate increased genetic stability of domain V of nOPV2 relative to mOPV2, with significantly lower neurovirulence of shed nOPV2 virus compared with shed mOPV2. While this vaccine is currently being deployed under an emergency use listing, the data on the genetic stability of nOPV2 will support further regulatory and policy decision-making regarding use of nOPV2 in outbreak responses. Bill & Melinda Gates Foundation.