Distribution of rotavirus strains and strain-specific effectiveness of the rotavirus vaccine after its introduction: a systematic review and meta-analysis

Background Concerns exist about whether monovalent (RV1) and pentavalent (RV5) rotavirus vaccines provide adequate protection against diverse strains and whether vaccine introduction will lead to selective pressure. We aimed to investigate the distribution of rotavirus strains and strain-specific rotavirus vaccine effectiveness after vaccine introduction. Methods We did a systematic review of published work to assess the strain-specific effectiveness of RV1 and RV5 rotavirus vaccines. We classified strains as homotypic, partly heterotypic, and fully heterotypic based on the amount of antigen-matching between strain and vaccine. When studies reported vaccine effectiveness against single antigens (G-type or P-type), we categorised them as either single-antigen vaccine type or single-antigen non-vaccine type. Our primary outcome was strain-specific vaccine effectiveness, comparing effectiveness of homotypic strains with fully or partly heterotypic strains. A secondary outcome was the prevalence of rotavirus strains after vaccine introduction. We estimated pooled odds ratios using random-effect regression models, stratified by country income level and vaccine type, and tested for differences in strain-specific vaccine effectiveness. We assessed strain distribution trends from surveillance reports. Findings In high-income countries, RV1 pooled vaccine effectiveness was 94% (95% CI 80-98) against homotypic strains, 71% (39-86) against partly heterotypic strains, and 87% (76-93) against fully heterotypic strains. In middle-income settings, respective pooled data were 59% (36-73), 72% (58-81), and 47% (28-61). In high-income countries, RV5 vaccine effectiveness was 83% (78-87) against homotypic strains, 82% (70-89) against single-antigen vaccine type strains, 82% (70-89) against partly heterotypic strains, and 75% (47-88) against single-antigen non-vaccine type strains. In middle-income settings, RV5 vaccine effectiveness was 70% (58-78) against single-antigen vaccine type strains, 37% (10-56) against partly heterotypic strains, and 87% (38-97) against single-antigen non-vaccine type strains. No difference was noted in vaccine effectiveness for either RV1 or RV5 in any setting (all p>0-05). Prevalent strains in countries using RV1 were G2P[4] (2198 of 4428,50%) and G1P[8] (953,22%), and those in countries using RV5 were G1P[8] (1280 of 3875, 33%) and G2P[4] (1169,30%). Sustained predominance of a single strain was not recorded. Interpretation RV1 and RV5 exert similar effectiveness against homotypic and heterotypic rotavirus strains. Persistence of specific strains was not recorded, suggesting vaccine-induced selective pressure did not occur. Expansion of rotavirus surveillance efforts to low-income countries and ongoing surveillance are crucial to identify emergence of new strains and to assess strain-specific vaccine effectiveness in various settings.