We did a randomised, controlled, open-label, non-inferiority trial with healthy, full-term (>2·5 kg birthweight) infants aged 8 weeks (± 7 days) at six well-child clinics in Santiago, Chile. We used supplied lists to randomly assign infants (1:1:1) to receive three polio vaccinations (IPV by injection or bOPV as oral drops) at age 8, 16, and 24 weeks in one of three sequential schedules: IPV–bOPV–bOPV, IPV–IPV–bOPV, or IPV–IPV–IPV. We did the randomisation with blocks of 12 stratified by study site. All analyses were done in a masked manner. Co-primary outcomes were non-inferiority of the bOPV-containing schedules compared with the all-IPV schedule for seroconversion (within a 10% margin) and antibody titres (within two-thirds log2 titres) to poliovirus serotypes 1 and 3 at age 28 weeks, analysed in the per-protocol population. Secondary outcomes were seroconversion and titres to serotype 2 and faecal shedding for 4 weeks after a monovalent OPV type 2 challenge at age 28 weeks. Safety analyses were done in the intention-to-treat population. This trial is registered with ClinicalTrials.gov, number NCT01841671, and is closed to new participants.
Between April 25 and August 1, 2013, we assigned 570 infants to treatment: 190 to IPV–bOPV–bOPV, 192 to IPV–IPV–bOPV, and 188 to IPV–IPV–IPV. 564 (99%) were vaccinated and included in the intention-to-treat cohort, and 537 (94%) in the per-protocol analyses. In the IPV–bOPV–bOPV, IPV–IPV–bOPV, and IPV–IPV–IPV groups, respectively, the proportions of children with seroconversion to type 1 poliovirus were 166 (98·8%) of 168, 95% CI 95·8–99·7; 178 (100%), 97·9–100·0; and 175 (100%), 97·9–100·0. Proportions with seroconvsion to type 3 poliovirus were 163 (98·2%) of 166, 94·8–99·4; 177 (100%), 97·9–100·0, and 172 (98·9%) of 174, 95·9–99·7. Non-inferiority was thus shown for the bOPV-containing schedules compared with the all-IPV schedule, with no significant differences between groups. In the IPV–bOPV–bOPV, IPV–IPV–bOPV, and IPV–IPV–IPV groups, respectively, the proportions of children with seroprotective antibody titres to type 1 poliovirus were 168 (98·8%) of 170, 95% CI 95·8–99·7; 181 (100%), 97·9–100·0; and 177 (100%), 97·9–100·0. Proportions to type 3 poliovirus were 166 (98·2%) of 169, 94·9–99·4; 180 (100%), 97·9–100·0; and 174 (98·9%) of 176, 96·0–99·7. Non-inferiority comparisons could not be done for this outcome because median titres for the groups receiving OPV were greater than the assay's upper limit of detection (log2 titres >10·5). The proportions of children seroconverting to type 2 poliovirus in the IPV–bOPV–bOPV, IPV–IPV–bOPV, and IPV–IPV–IPV groups, respectively, were 130 (77·4%) of 168, 95% CI 70·5–83·0; 169 (96·0%) of 176, 92·0–98·0; and 175 (100%), 97·8–100. IPV–bOPV schedules resulted in almost a 0·3 log reduction of type 2 faecal shedding compared with the IPV-only schedule. No participants died during the trial; 81 serious adverse events were reported, of which one was thought to be possibly vaccine-related (intestinal intussusception).
Seroconversion rates against polioviruses types 1 and 3 were non-inferior in sequential schedules containing IPV and bOPV, compared with an all-IPV schedule, and proportions of infants with protective antibodies were high after all three schedules. One or two doses of bOPV after IPV boosted intestinal immunity for poliovirus type 2, suggesting possible cross protection. Additionally, there was evidence of humoral priming for type 2 from one dose of IPV. Our findings could give policy makers flexibility when choosing a vaccination schedule, especially when trying to eliminate vaccine-associated and vaccine-derived poliomyelitis.
Bill & Melinda Gates Foundation.