Rotavirus is the main cause of severe acute gastroenteritis in children in Africa. Monovalent human rotavirus vaccine (RV1) was added into Malawi's infant immunisation schedule on Oct 29, 2012. We aimed to assess the impact and effectiveness of RV1 on rotavirus gastroenteritis in the 2 years after introduction. From Jan 1, 2012, to June 30, 2014, we recruited children younger than 5 years who were admitted into Queen Elizabeth Central Hospital, Blantyre, Malawi, with acute gastroenteritis. We assessed stool samples from these children for presence of rotavirus with use of ELISA and we genotyped rotaviruses with use of RT-PCR. We compared rotavirus detection rates in stool samples and incidence of hospital admittance for rotavirus in children from Jan 1 to June 30, in the year before vaccination (2012) with the same months in the 2 years after vaccination was introduced (2013 and 2014). In the case-control portion of our study, we recruited eligible rotavirus-positive children from the surveillance platform and calculated vaccine effectiveness (one minus the odds ratio of vaccination) by comparing infants with rotavirus gastroenteritis with infants who tested negative for rotavirus, and with community age-matched and neighbourhood-matched controls. We enrolled 1431 children, from whom we obtained 1417 stool samples (99%). We detected rotavirus in 79 of 157 infants (50%) before the vaccine, compared with 57 of 219 (40%) and 52 of 170 (31%) in successive calendar years after vaccine introduction (p=0·0002). In the first half of 2012, incidence of rotavirus hospital admission was 269 per 100,000 infants compared with 284 in the same months of 2013 (rise of 5·8%, 95% CI -23·1 to 45·4; p=0·73) and 153 in these months in 2014 (a reduction from the prevaccine period of 43·2%, 18·0-60·7; p=0·003). We recruited 118 vaccine-eligible rotavirus cases (median age 8·9 months; IQR 6·6-11·1), 317 rotavirus-test-negative controls (9·4 months; 6·9-11·9), and 380 community controls (8·8 months; 6·5-11·1). Vaccine effectiveness for two doses of RV1 in rotavirus-negative individuals was 64% (95% CI 24-83) and community controls was 63% (23-83). The point estimate of effectiveness was higher against genotype G1 than against G2 and G12. Routine use of RV1 reduced hospital admissions for several genotypes of rotavirus in children younger than 5 years, especially in infants younger than 1 year. Our data support introduction of rotavirus vaccination at the WHO recommended schedule, with continuing surveillance in high-mortality countries. Wellcome Trust, GlaxoSmithKline Biologicals.

Background. Rotavirus vaccines have been introduced in many low-income African countries including Malawi in 2012. Despite early evidence of vaccine impact, determining persistence of protection beyond infancy, the utility of the vaccine against specific rotavirus genotypes, and effectiveness in vulnerable subgroups is important. Methods. We compared rotavirus prevalence in diarrheal stool and hospitalization incidence before and following rotavirus vaccine introduction in Malawi. Using case-control analysis, we derived vaccine effectiveness (VE) in the second year of life and for human immunodeficiency virus (HIV)–exposed and stunted children. Results. Rotavirus prevalence declined concurrent with increasing vaccine coverage, and in 2015 was 24% compared with prevaccine mean baseline in 1997–2011 of 32%. Since vaccine introduction, population rotavirus hospitalization incidence declined in infants by 54.2% (95% confidence interval [CI], 32.8–68.8), but did not fall in older children. Comparing 241 rotavirus cases with 692 test-negative controls, VE was 70.6% (95% CI, 33.6%–87.0%) and 31.7% (95% CI, −140.6% to 80.6%) in the first and second year of life, respectively, whereas mean age of rotavirus cases increased from 9.3 to 11.8 months. Despite higher VE against G1P[8] than against other genotypes, no resurgence of nonvaccine genotypes has occurred. VE did not differ significantly by nutritional status (78.1% [95% CI, 5.6%–94.9%] in 257 well-nourished and 27.8% [95% CI, −99.5% to 73.9%] in 205 stunted children; P = .12), or by HIV exposure (60.5% [95% CI, 13.3%–82.0%] in 745 HIV-unexposed and 42.2% [95% CI, −106.9% to 83.8%] in 174 exposed children; P = .91). Conclusions. Rotavirus vaccination in Malawi has resulted in reductions in disease burden in infants <12 months, but not in older children. Despite differences in genotype-specific VE, no genotype has emerged to suggest vaccine escape. VE was not demonstrably affected by HIV exposure or stunting.

Identifying risk factors for impaired oral rotavirus vaccine (ORV) efficacy in low-income countries may lead to improvements in vaccine design and delivery. In this prospective cohort study, we measure maternal rotavirus antibodies, environmental enteric dysfunction (EED), and bacterial gut microbiota development among infants receiving two doses of Rotarix in India (n = 307), Malawi (n = 119), and the UK (n = 60), using standardised methods across cohorts. We observe ORV shedding and seroconversion rates to be significantly lower in Malawi and India than the UK. Maternal rotavirus-specific antibodies in serum and breastmilk are negatively correlated with ORV response in India and Malawi, mediated partly by a reduction in ORV shedding. In the UK, ORV shedding is not inhibited despite comparable maternal antibody levels to the other cohorts. In both India and Malawi, increased microbiota diversity is negatively correlated with ORV immunogenicity, suggesting that high early-life microbial exposure may contribute to impaired vaccine efficacy.

To combat the high burden of rotavirus gastroenteritis, multiple African countries have introduced rotavirus vaccines into their childhood immunization programs. Malawi incorporated a G1P[8] rotavirus vaccine (Rotarix) into its immunization schedule in 2012. Utilizing a surveillance platform of hospitalized rotavirus gastroenteritis cases, we examined the phylodynamics of G1P[8] rotavirus strains that circulated in Malawi before (1998 to 2012) and after (2013 to 2014) vaccine introduction. Analysis of whole genomes obtained through next-generation sequencing revealed that all randomly selected prevaccine G1P[8] strains sequenced ( n = 32) possessed a Wa-like genetic constellation, whereas postvaccine G1P[8] strains ( n = 18) had a DS-1-like constellation. Phylodynamic analyses indicated that postvaccine G1P[8] strains emerged through reassortment events between human Wa- and DS-1-like rotaviruses that circulated in Malawi from the 1990s and hence were classified as atypical DS-1-like reassortants. The time to the most recent common ancestor for G1P[8] strains was from 1981 to 1994; their evolutionary rates ranged from 9.7 × 10 ⁻⁴ to 4.1 × 10 ⁻³ nucleotide substitutions/site/year. Three distinct G1P[8] lineages chronologically replaced each other between 1998 and 2014. Genetic drift was the likely driver for lineage turnover in 2005, whereas replacement in 2013 was due to reassortment. Amino acid substitution within the outer glycoprotein VP7 of G1P[8] strains had no impact on the structural conformation of the antigenic regions, suggesting that it is unlikely that they would affect recognition by vaccine-induced neutralizing antibodies. While the emergence of DS-1-like G1P[8] rotavirus reassortants in Malawi was therefore likely due to natural genotype variation, vaccine effectiveness against such strains needs careful evaluation. IMPORTANCE The error-prone RNA-dependent RNA polymerase and the segmented RNA genome predispose rotaviruses to genetic mutation and genome reassortment, respectively. These evolutionary mechanisms generate novel strains and have the potential to lead to the emergence of vaccine escape mutants. While multiple African countries have introduced a rotavirus vaccine, there are few data describing the evolution of rotaviruses that circulated before and after vaccine introduction. We report the emergence of atypical DS-1-like G1P[8] strains during the postvaccine era in Malawi. Three distinct G1P[8] lineages circulated chronologically from 1998 to 2014; mutation and reassortment drove lineage turnover in 2005 and 2013, respectively. Amino acid substitutions within the outer capsid VP7 glycoprotein did not affect the structural conformation of mapped antigenic sites, suggesting a limited effect on the recognition of G1-specific vaccine-derived antibodies. The genes that constitute the remaining genetic backbone may play important roles in immune evasion, and vaccine effectiveness against such atypical strains needs careful evaluation.