Sapovirus Infections in an Australian Community-Based Healthy Birth Cohort During the First 2 Years of Life

METHODS

The ORChID study was a prospective community-based, healthy birth cohort based in Brisbane, Australia (ClinicalTrials.gov: NCT01304914) [20]. Recruitment occurred over 2 years, from September 2010 to October 2012, and data were collected until October 2014. Participants were enrolled from birth in 2 hospitals (1 public and 1 privately funded). The Human Research Ethics Committees of the Children's Health Queensland Hospital and Health Service, the Royal Brisbane and Women's Hospital (HREC/10/QRBW/125), and The University of Queensland (2010000820) approved the study.

RESULTS

Overall, 158 children provided 11 124 swabs before their second birthday, and 154 returned 87 547 symptom diary-days of observation (78% of the possible maximum) (Supplementary Figure 2). There were 10 820 swabs matched to 82 104 diary-days (Supplementary Figure 3). Table 2 shows that 65% of participants were first-born, 63% had university-educated mothers, and 56% came from the top quartile of household incomes. In total, 129 of 131 (99%) children received 3 doses of the RotaTeq rotavirus vaccine (Merck & Co, Whitehouse Station, New Jersey).

DISCUSSION

This is the first community-based longitudinal birth cohort study to describe the epidemiology of sapovirus infections during the first 2 years of life in an HIC with an established rotavirus vaccine program [28]. The IR we observed (0.89 infections per child-year) was similar to estimates from a recent Peruvian birth cohort study (0.94 infections per child-year) [17]. Although our IR for symptomatic episodes (0.26 cases per child-year) exceeded that reported from Nicaragua (0.13 cases per child-year) [29] for the same age group, it was lower than reported from Peru (0.52 cases per child-year) [17] and the MAL-ED study (0.47 cases per child-year) [16].

Unlike other birth cohort studies, the ORChID study maximized the chances of capturing sapovirus infections by parents collecting regular weekly stool samples. In the Nicaraguan study of 444 children, fieldworkers collected monthly routine stool samples and at times of an AGE illness, which included vomiting and/or diarrhea [29]. Similarly in the MAL-ED study, fieldworkers visited 1715 children twice weekly and collected stool samples whenever there were diarrheal symptoms, as well as routinely each month from ages 1 to 12 months, then every 3 months until 24 months of age [16]. In Peru, 100 children were followed until 2 years of age and stool samples collected during diarrheal episodes and every 3 months [17]. In a subset of 47 children with 67 sapovirus infections from this cohort, 440 weekly stool samples were collected and the median duration of sapovirus shedding was 18.5 days (IQR, 12.5–23.5 days), similar to our findings. It is possible these LMIC studies may not have detected all symptomatic episodes when isolated vomiting did not trigger a stool sample. Similarly, with virus shedding duration usually ≤4 weeks, asymptomatic infections may also have been missed as all 3 LMIC cohorts reported asymptomatic sapovirus infections in less than half of their identified cases [16, 17, 29].

Consistent with these birth-cohort studies [16, 17, 29], we found that children aged ≥6 months were at greatest risk of sapovirus infections, possibly reflecting the protective role of passively acquired maternal antibodies early in life [30]. As with the MAL-ED study, we found that the IR for symptomatic episodes peaked between 6 and 17 months of age [16]. In addition to age, attending formal childcare was an independent risk factor for sapovirus infections. This supports a study of Danish childcare centers that found 21% of attendees shedding sapoviruses, suggesting that children attending childcare centers may act as reservoirs for virus transmission [19]. Although sapoviruses were detected year-round, numbers in the ORChID cohort peaked each fall season. Such distinct seasonality is not a universal finding. Seasonal peaks have been reported in Spain in the fall and late winter/early spring [31], in winter and spring in the Netherlands [32], and in summer in Japan [33], but with little evidence of seasonality in the United States [8, 9].

Although the possibility for virus–virus interactions is widely recognized, their exact nature and consequences are still not fully understood. Coinfections are more frequently thought of as viruses from different viral families producing simultaneous infection [34]. In agreement with recent studies [16, 34], our data found a high risk of coinfection with different enteric viruses in children infected with sapovirus. Among these enteric viruses, rotavirus posed the highest risk of sapovirus coinfection, which is consistent with the MAL-ED study [16]. In contrast, we observed a negative association between sapovirus and norovirus infections. A similar trend for this negative association was reported in an AGE outbreak in Finland, where 72% of the collected samples were positive for at least 2 enteric viruses; however, there were no cases where >1 type of calicivirus was present [35].

In the ORChID cohort symptomatic sapovirus episodes were detected in just 30% of all infections. In our study, 72% of the symptomatic episodes included vomiting, while 48% had diarrhea. In several community-based studies, 23%–90% of symptomatic episodes have vomiting as a symptom [18, 29, 36, 37] compared to the 72% we observed. This variation may be due to differences in how symptomatic cases were identified, as vomiting as a sole symptom was not always adopted as a diagnostic criterion for AGE episodes. It seems that young children with sapovirus and norovirus are more likely to have a symptom profile of a vomiting-only illness than those infected with other enteric viruses [8, 27, 37, 38]. This is supported by our results where we found that 52% of symptomatic sapovirus episodes were associated with vomiting as a sole symptom.

Children with symptomatic episodes had almost 10-fold higher viral loads [22] and shed sapoviruses for about 4 days longer than those with asymptomatic infections. Nevertheless, those with asymptomatic infections can also excrete large quantities of virus (10⁶–10⁹ viral copies per gram of stool) [39], and children with asymptomatic infections might contribute to transmission and outbreaks in the community [19, 39].

Overall, 43% of symptomatic sapovirus episodes resulted in healthcare visits and 7% of all symptomatic cases were defined as severe. However, those with prior infections had a 54% lower risk of developing symptoms than those experiencing a primary infection and shed the virus for slightly shorter periods. Nonetheless, if symptoms occurred there was no difference in healthcare encounters between children with primary and subsequent infections.

An outpatient study of AGE in Spain reported highest case numbers of sapovirus at the extremes of age, suggesting that infections in childhood may provide some protection against infections later in life, only to decline with aging [31]. Although in our cohort reinfections were common, our data suggest that children with at least 2 sapovirus infections have a 53% reduction in the hazard of subsequent infection. The Peruvian birth cohort study sequenced almost two-thirds of their sapovirus-positive samples [17]. In a subset of children with multiple infections, 51% were reinfected with strains from the same genogroup, while just 9% were reinfected with the same genotype, suggesting predominantly genotype-specific protective immunity. A recent Nicaraguan birth cohort study found a 69% reduction in hazard against repeat symptomatic episodes following a first symptomatic sapovirus episode [40], while 2 sapovirus infections reduced the hazard of further infections by 31% in MAL-ED participants [16]. More research is needed to determine the extent to which reinfection with a specific genotype provides cross-protection against symptomatic infection with other genotypes or genogroups.

The strengths of the ORChID study are its prospective design, which allowed us to investigate the epidemiology of both asymptomatic and symptomatic sapovirus infections using high-density symptom recording and weekly stool samples. We also included vomiting-only cases in our definition of a symptomatic episode. However, there are important limitations. We did not differentiate between genogroups or genotypes of sapoviruses. Hence, we were unable to investigate whether some sapovirus strains provide homotypic or heterotypic protection in children. We also used Ct values, which only provide semi-quantitative estimates of viral loads. The overall IR estimate may be slightly underestimated as swabs were less likely to be returned as the child aged, which is also when sapovirus incidence was higher. Finally, as often occurs with intense studies of this nature, generalizability of our results may be limited to similar household income, educational, and climatic settings.

In conclusion, in the first 2 years of life there is a high incidence of sapovirus infections among Australian children aged >6 months. Sapoviruses were associated with childcare and fall season peaks. Only 30% were symptomatic of whom 52% had vomiting as their only symptom, while 43% of symptomatic cases led to healthcare visits. Repeat infections are common, but cumulatively reduce the risk of further infections. Although children with asymptomatic infections shed lower virus loads for shorter periods, they may be important reservoirs for sapovirus transmission. Our results suggest that sapoviruses, like noroviruses from the same cohort [27], are leading causes of gastrointestinal infections in Australian children <2 years of age. To maximize capturing symptomatic cases, future studies should include vomiting-only episodes as part of their case definitions. Preventive measures like those recommended for norovirus of hand and environmental hygiene and avoiding contact with ill people may help decrease disease transmission. In the future, vaccines against sapovirus, either alone or combined with norovirus, could contribute to reducing the overall burden of childhood AGE, potentially providing indirect protection in older age groups. Estimates of sapovirus incidence in communities, particularly among children, remain critical for future vaccine development and public health policies.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments. A. E. is a recipient of Higher Degree Research Scholarships from the Australian Government Research Training Program.

Data sharing. The data that support the findings of this study are available on reasonable request from the corresponding author. The data are not publicly available because they include information that could compromise the privacy of research participants.

Financial support. This study was supported by an Australian National Health and Medical Research Council (NHMRC) project grant (grant number GNT615700) and a program grant from the Children's Hospital Foundation Queensland (CHFQ) (grant number 5006) under Children's Hospital Foundation, Australia. S. B. L. was the recipient of an NHMRC Early Career Fellowship and a CHFQ Mid-Career Fellowship. These funding bodies had no role in study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.

Supplementary Material