Use of Surveillance Data to Assess the Impact of Vaccination on Circulating Rotavirus Strains

(See the Original Article by Wylie, Weinstock, and Storch on pages e84–e89.) Circulating rotavirus strains in a given location show natural secular variation with sometimes dramatic and rapid changes over successive rotavirus seasons [ 1 , 2 ]. This temporal variation makes it particularly challenging to determine whether any observed change in prevalent strains after implementation of rotavirus vaccination merely represents naturally occurring strain variation or instead the more concerning phenomenon of selection of strains against which vaccine effectiveness is lower (ie, selective pressure) [ 3 ]. In the current issue of the Journal of Pediatric Infectious Disease Society , Wylie et al [ 4 ] report the emergence during the 2012–2013 season of G12, a previously unusual rotavirus strain in the St. Louis area. The authors raise the question of whether this G12 emergence is related to the use of rotavirus vaccine implemented in 2006 in this population? To address this question, it is important to first consider the evidence from prelicensure trials on strain-specific protection with rotavirus vaccines. The 2 rotavirus vaccines recommended by the Advisory Committee on Immunization Practices for use in US children differ in their composition: RotaTeq (RV5; RotaTeq, Merck and Co., Whitehouse Station, NJ) is a pentavalent (G1, G2, G3, G4, P[ 8 ]) human-bovine reassortant vaccine, whereas Rotarix (RV1; Rotarix, GSK Biologicals, Rixensart, Belgium) is a monovalent (G1P[ 8 ]) vaccine derived from a single attenuated human rotavirus strain [ 5 ]. In clinical trials, both vaccines protected against disease caused by a range of rotavirus strains that were homotypic (with both the G- and P-types included in vaccine), fully heterotypic (neither G- nor P-type included in vaccine), or partially heterotypic (only 1 of the G- or P-type included in vaccine). Although the point estimate of efficacy of RV1 was lower against fully heterotypic G2P[ 4 ] strains (41%; 95% confidence interval [CI], 79%–82%) compared with other strains (efficacy range, 87%–92%) in the pivotal prelicensure trial in Latin America, the number of G2P[ 4 ] cases was small, and thus firm conclusions could not be drawn [ 6 ]. Reassuringly, a later RV1 efficacy trial in Europe conducted during a period when G2P[ 4 ] strains were more prevalent showed good efficacy (85%; 95% CI, 24%–98%) against disease cause by this strain [ 7 ]. A subsequent trial of RV1 in Africa also reaffirmed that the vaccine provided good protection against a range of strains (G8, G9, G12) that are not included in the vaccine [ 8 ]. In addition, a recently published clinical trial of the Indian-made monovalent human-bovine (116E) rotavirus vaccine based on an unusual G9P[ 11 ] neonatal rotavirus strain showed good protection against a range of heterotypic strains [ 9 ]. The RV5 was also shown to be effective against heterotypic strains in clinical trials conducted in various settings [ 10 , 11 ]. Collectively, these data from various vaccine trials underscore the broad cross-protection seen with rotavirus vaccines and suggest that immunity to rotavirus is likely polygenic.