Resurgence of Rotavirus Genotype G12 in St. Louis During the 2014–2015 Rotavirus Season

Abstract

Background

Rotaviruses are a leading cause of gastroenteritis. Rotavirus vaccination has dramatically reduced rotavirus occurrence; however, we have noticed mild to moderate recurrences in the St. Louis area in alternate years. In 2013, we found rotavirus genotype G12 to be the dominant strain in the St. Louis region. In this study, we again determined the distribution of genotypes and ascertained vaccine history in patients infected with rotavirus G12 during the 2014–15 season.

Methods

Samples submitted to the St. Louis Children’s Hospital Microbiology Laboratory were tested for rotavirus using an antigen assay. We determined the VP7 genotype using amplicon sequence analysis. We determined genome sequences using high-throughput sequencing. We evaluated rotavirus immunization records when available.

Results

Of 30 typed viruses from 2014–15, 29 were G12 (97%). Whole-genome sequencing revealed few changes from G12 viruses analyzed in 2012–13. VP4 and VP7 sequences were >99% identical to previously sequenced G12 strains from St. Louis, and immune epitopes were conserved. Vaccination histories were available from 17 patients. Of these, 4 had been vaccinated, 3 had received incomplete vaccination or had a vaccine history that could not be confirmed, and 10 had not been vaccinated.

Conclusions

G12 re-emerged as the predominant rotavirus genotype in 2014–15, comprising a higher percentage of cases than in 2012–13. The majority of patients with G12 and available vaccination histories were unvaccinated. There was no genomic evidence to indicate that the G12 strains in St. Louis had evolved to escape vaccine protection. Our work emphasizes the need for continued surveillance.

Rotaviruses are a leading cause of severe acute gastroenteritis and dehydration in children and infants. The World Health Organization estimated that globally rotaviruses caused 527000 deaths per year in children under the age of 5 annually before the introduction of vaccination programs in 2006 [1]. Vaccination has been highly efficacious in reducing the burden of disease from rotavirus infection [2–6], with a more recent estimate from the World Health Organization that approximates 215000 annual rotavirus-related deaths [7]. Vaccines have been shown to provide heterotypic protection against antigenically diverse rotaviruses that are not specifically included in the vaccine strains [8–10]. Even with this success, the emergence of new genotypes and re-emergence of old threats necessitates continued broad global surveillance of rotavirus genotypes [11].

Rotaviruses, from the family Reoviridae, have double-stranded, ribonucleic acid genomes with 11 segments. Rotaviruses can be classified based on the sequence of the VP4 gene (the P type) and the VP7 gene (the G type). These genes encode the attachment protein and glycoprotein, respectively, which contain the primary epitopes that elicit an immune response [12]. Antigenic epitopes have also been predicted on other proteins [13–15], although their roles in enhancing or stimulating antiviral immunity are not entirely clear. Globally, the most common strains have been types G1P [8], G2P [4], G3P [8], G4P [8], and G9P [8], although there are more than 70 G-P combinations [11, 16]. The RotaTeq vaccine (Merck and Co., Inc) was licensed by the US Food and Drug Administration and approved by the US Advisory Committee for Immunization Practices for release in the United States in 2006, followed by Rotarix (GlaxoSmithKline Biologicals), which was released in 2008. RotaTeq (also called RV5) is a pentavalent vaccine. It includes 5 human-bovine reassortant rotaviruses, containing the antigenic segments from human rotaviruses that were most common before vaccine development (VP7 types G1, G2, G3, and G4 and VP4 type P [8]). Rotarix (RV1) is a live-attenuated monovalent vaccine consisting of the common human G1P [8] type.

VP7 type G12 has emerged as an increasingly common type globally. Previously, G12 appeared to be limited to Southeast Asia and Africa [17]; however, in recent years, G12 has been reported from multiple locations including Europe [18, 19], South America [20], Central America [21], and Australia [22]. An outbreak occurred in Rochester, New York in 2007 but did not persist during the succeeding season [23], and G12 was reported as the dominant type in 2010 in a small study from Chicago [24] and in a food-borne gastroenteritis outbreak near Chicago in 2013 [25]. A recent study from the New Vaccine Surveillance network reported that 72% of rotaviruses detected from 7 centers in the United States between December 1, 2011 and March 31, 2013 were G12 [26, 27]. In 2013, we documented the emergence of G12 in St. Louis, Missouri, finding G12 in 24 of 28 typed samples (86%) [28]. During the following 2 years, we have continued to monitor the detection of rotavirus at St. Louis Children’s Hospital (SLCH) to track the natural history of rotavirus types in the St. Louis area.

METHODS

The work was carried out under protocols approved by the Washington University Human Research Protection Office. Rotavirus testing was conducted on stool samples submitted to the Microbiology Laboratory at SLCH using the X/pect Rotavirus rapid antigen test (Remel, Lenexa, KS). Antigen testing was performed on 165 total samples, 39 (24%) of which were positive. Typing was performed on 30 samples for which residual material was available after clinical testing had been completed, and no selection criteria were applied. VP7 genotyping to determine the G type was performed using reverse transcriptase-polymerase chain reaction followed by nucleotide sequencing of the amplification product [29]. Sequences were compared and typed using the Basic Local Alignment Search Tool (BLAST) [30].

Rotavirus genome sequencing was attempted on total nucleic acid extracted from 24 samples (Supplemental Table 1). Samples were chosen to represent the full timespan of the season; when 2 samples were collected on the same day, the first sample identifier was chosen after alphanumeric sorting. A description of the methods used for genomic sequencing is included in the Supplemental Methods file. Sequence data were submitted to the Sequence Read Archive under Bioproject ID PRJNA221644, Biosample IDs SAMN05660249-05660266. We ascertained vaccine status by contacting guardians and medical providers.

RESULTS

The number of positive stool samples and percentage positive for samples collected and processed in the Microbiology Laboratory at SLCH for rotavirus seasons 2005 through 2016 are shown in Figure 1. The rotavirus season is defined as December–May. There were only 3 positive samples during the 2013–14 season. Genotyping was successful on only one, which was G2. Of 30 typed rotavirus-positive samples from the 2014–15 season, 29 (97%) were G12 and 1 was G3. Rotavirus-positive patients resided throughout the St. Louis metropolitan area without evidence of geographical clustering (Figure 2). The ages of the 39 rotavirus-positive patients are shown in Figure 3. The median patient age was 2.1 years (range 2 weeks–59 years), compared with 2.5 years (range, 3 weeks–10 years) for all rotavirus-positive patients during the 2012–13 season. There was no statistical difference in the ages of the patients between the 2 seasons (Wilcoxon-Mann-Whitney test). Vaccine histories were obtained for 17 of the rotavirus-positive patients. Of these, 10 had not received the vaccine, 3 had begun the vaccination process with RotaTeq but had not received all doses (n = 2) or receipt of all doses could not be confirmed (n = 1), and 4 had received the full series, including 2 who had received Rotarix and 2 who had received RotaTeq. Of note, all 6 of the patients under the age of 2 years for whom vaccine information was available and who were old enough to have received the vaccine when they became ill were fully (n = 4) or incompletely vaccinated (n = 2). In contrast, 8 of the 9 patients over the age of 2 years had not been vaccinated, and the remaining patient was the one for whom receipt of the full series could not be confirmed. Vaccine uptake during 2014 in the states of Missouri and Illinois for children 19–35 months was estimated to be 74.4% and 73.7%, respectively, by the Center for Disease Control’s National Immunization Survey [31].

Figure 1.

Rotavirus testing results. The x-axis shows the rotavirus season. The left y-axis shows the number of postive rotavirus tests in the clinical laboratory, indicated by the bars. The right y-axis shows the percentage of tests that were positive for rotavirus, indicated by the line.

Figure 2.

Location of residence of patients with rotavirus during the 2014–15 rotavirus season. Zip codes were used to map patients’ residence using BatchGeo (http://batchgeo.com/). Residence was available for 37 cases. The larger map shows the distribution of cases through Missouri, Illinois, and Oklahoma. The St. Louis region is magnified in the inset to show the distribution throughout the metropolitan area.

Figure 3.

Age distribution of patients in the 2014–15 rotavirus season in St. Louis. Age are displayed in the histogram. The x-axis shows age in years. The y-axis indicates the number of subjects.

Rotavirus genome sequencing was attempted for 24 samples. Complete sequence coverage of all 12 open reading frames (11 genome segments) was obtained for viruses in 11 samples. Each of these was typed as G12-P [8]-I1-R1-C1-M1-A1-N1-T1-E1-H1, consistent with the Wa-like genogroup 1 [32], the same genogroup that we observed in 2013 [28]. Viruses from 2014–15 were >99% identical to each other at the nucleotide level and to a representative strain from the 2012–13 St. Louis outbreak (RVAA-9), indicating minimal evolution in the 2 years between 2 biennial resurgences (Figure 4). We observed no predicted amino acid sequence changes between the 2012–13 and the 2014–15 seasons in the sequences encoding the well defined antigenic epitopes in the VP4 and VP7 proteins [33–35], except for one sample (RVAA-17, whose genome was not fully sequenced), which had a D97N change in the VP7 protein sequence (epitopes examined shown in Supplemental Table 2). We also examined predicted linear antigenic epitopes of other proteins, which were found in the Immune Epitope Database and Analysis Resource (IEDAR), shown in Supplemental Table 3. Of 77 predicted epitope sequences, 62 were completely conserved in the St. Louis viruses from 2014–15. Although the other 15 epitope sequences in the 2014–15 St. Louis strains varied from the sequences in the IEDAR database, we saw no novel changes in any of the predicted epitopes in the 2014–15 viruses compared with 2012–2013 St. Louis viruses.

Figure 4.

VP7 phylogenetic tree. The St. Louis VP7 sequences were compared with other G12 VP7 sequences using the maximum likelihood method. The strain L26, a G12 virus from 1987, is included for reference as a more distantly related virus. The 2014–15 St. Louis viruses are marked with black triangles to the right. The 2012–13 St. Louis virus is marked with a black circle to the right. The tree scale shown represents the mean number of substitutions per position. Bootstrap confidence values greater than 70% from 1000 bootstrap repetitions are shown at the nodes.

DISCUSSION

During the 2014–15 rotavirus season in St. Louis, the prevailing VP7 genotype was again G12, accounting for a remarkable 97% of genotyped viruses. The natural history of G12 in St. Louis can be traced back at least as far as the 2010–11 rotavirus season, when G12 constituted 15% of rotavirus-positive samples [28]. After minor activity during the 2011–12 season, a resurgence of rotavirus occurred in St. Louis in the 2012–13 season and 86% of typed viruses were G12 [28]. Another resurgence with G12 becoming increasingly dominant occurred during the 2014–15, continuing a biennial pattern of rotavirus occurrence that has also been noted elsewhere [36, 37].

Globally, G12 constituted only a small percentage of rotavirus cases from 2003 to 2007, representing 1.3% of all strains, whereas G1 at 42.9% continued to be the most common genotype [17]. Since that time, there have been reports of the emergence of G12 as a dominant genotype in many locations [18–21]. Reports from the Center for Disease Control’s New Vaccine Surveillance Network indicate that G12 accounted for 10% of rotavirus-positive samples collected during the 2009–2011 seasons from 7 sites in the United States [9] and 72% in a report of samples collected from 7 sites during the 2011–2013 seasons [38]. These results suggest that the emergence of rotavirus G12 is not limited to the St. Louis area and may be widely present in the United States. For the second time in 3 years, G12 has reappeared and become increasingly dominant in the greater St. Louis area.

It is also difficult to draw conclusions about the relationship between the emergence of rotavirus G12 and the widespread implementation of a rotavirus vaccination program in the United States The fact that G12 began to emerge before widespread use of the current rotavirus vaccines [39] suggests that vaccine utilization is not the sole explanation. Furthermore, rapid shifts in rotavirus genotype distribution have occurred in the past without apparent relationship to vaccine use [40, 41], suggesting that the emergence of G12 may be related to factors other than rotavirus vaccine. It is also possible that the vaccine has effectively controlled previously common genotypes, opening an ecologic niche for a new type such as G12. However, this explanation remains speculative at this time.

The emergence of a genotype different from those in licensed rotavirus vaccines raises the concern of decreased heterotypic protection. In this regard, it is reassuring that most of the children with G12 in this report for whom we could obtain vaccination history had not been vaccinated. There were a small number of apparent vaccine failures, 2 of which only occurred in children documented to have G12. The full genome sequence was determined for 1 of the 2 and was almost identical to the sequence of rotaviruses that we sequenced from the 2012–2013 outbreak, suggesting that the virus is not exhibiting rapid evolution, either in recognized antigenic epitopes or elsewhere in the genome.

This study has several limitations. The samples tested represent a convenience sample, and although we have no reason to suspect that the samples are not representative, there could be unsuspected biases in the children who were tested and for whom residual sample material was available. In addition, the samples came only from a single hospital; however, the catchment area of the hospital spans the entire St. Louis metropolitan area. The study was retrospective and the number of children for whom we were able to confirm vaccine histories was limited. Whole genome sequences were not obtained for all samples, although we have no reason to suspect the genomes that have been sequenced and analyzed are not representative of the population sampled. This study does not directly address whether the re-emergence of G12P [8] is due to vaccine pressure on strains included in or more similar to those represented in the vaccine strains. However, heterotypic protection against G12 strains has been demonstrated [9], and we do not observe genomic changes that suggest the 2014–15 G12P [8] viruses have evolved to escape the vaccine since their first appearance in St. Louis in the 2010–11 season.

CONCLUSIONS

In summary, rotavirus G12 has emerged as the dominant rotavirus in the St. Louis metropolitan area. As noted nationally [36, 37], rotavirus activity has exhibited a pattern of resurgence in alternate years. It is important to recognize that the strong downward trend in rotavirus incidence that began with implementation of a national vaccine program continues. Our detailed genomic analysis reveals conservation of the G12 viruses over 3 years. Most of the children with G12 were not immunized, suggesting that vaccine failure is not the explanation for the alternate year rotavirus resurgences. Nevertheless, continued virologic and genomic surveillance is of great importance.

Supplementary Data

Supplementary materials are available at Journal of the Pediatric Infectious Diseases Society online.

Notes

Acknowledgments. We thank Richard Buller and Stephanie Bledsoe from the Clinical Virology Laboratory at Children’s Hospital in St. Louis for providing samples and Brandi Herter and Maria Cannella for their assistance with the high-throughput sequencing and virus subtyping assays.

Financial support. This work was supported by institutional funds from the Washington University Department of Pediatrics. The Virus Pathogen Database and Analysis Resource, a publicly available resource used in this study, has been wholly funded with federal funds from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health and Human Services, under Contract No. HHSN272200900041C.

Potential conflicts of interest. All authors: No reported conflicts.

All authors have submitted the ICMJE Form for Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.