Vaccine schedules and the effect on humoral and intestinal immunity against poliovirus: a systematic review and network meta-analysis

The eradication of wild and vaccine-derived poliovirus requires the global withdrawal of oral poliovirus vaccines (OPVs) and replacement with inactivated poliovirus vaccines (IPVs). The first phase of this effort was the withdrawal of the serotype 2 vaccine in April 2016, with a switch from trivalent OPVs to bivalent OPVs. The aim of our study was to produce comparative estimates of humoral and intestinal mucosal immunity associated with different routine immunisation schedules. We did a random-effect meta-analysis with single proportions and a network meta-analysis in a Bayesian framework to synthesise direct and indirect data. We searched MEDLINE and the Cochrane Library Central Register of Controlled Trials for randomised controlled trials published from Jan 1, 1980, to Nov 1, 2018, comparing poliovirus immunisation schedules in a primary series. Only trials done outside western Europe or North America and without variation in age schedules (ie, age at administration of the vaccine) between study groups were included in the analyses, because trials in high-income settings differ in vaccine immunogenicity and schedules from other settings and to ensure consistency within the network of trials. Data were extracted directly from the published reports. We assessed seroconversion against poliovirus serotypes 1, 2, and 3, and intestinal immunity against serotype 2, measured by absence of shedding poliovirus after a challenge OPV dose. We identified 437 unique studies; of them, 17 studies with a maximum of 8279 evaluable infants were eligible for assessment of humoral immunity, and eight studies with 4254 infants were eligible for intestinal immunity. For serotype 2, there was low between-trial heterogeneity in the data (τ=0·05, 95% credible interval [CrI] 0·009–0·15) and the risk ratio (RR) of seroconversion after three doses of bivalent OPVs was 0·14 (95% CrI 0·11–0·17) compared with three doses of trivalent OPVs. The addition of one or two full doses of an IPV after a bivalent OPV schedule increased the RR to 0·85 (0·75–1·0) and 1·1 (0·98–1·4). However, the addition of an IPV to bivalent OPV schedules did not significantly increase intestinal immunity (0·33, 0·18–0·61), compared with trivalent OPVs alone. For serotypes 1 and 3, there was susbstantial inconsistency and between-trial heterogeneity between direct and indirect effects, so we only present pooled estmates on seroconversion, which were at least 80% for serotype 1 and at least 88% for serotype 3 fo