Argentina was one of the latest countries to adopt universal rotavirus vaccination in the Americas. However, rapid and successful results were observed regarding burden of diarrheal disease [14]. Following four years of this strategy, the prevalence of rotavirus remained low in children under 5 years of age in comparison with the previous study [14]. These findings oppose to other experiences that had described an increase in the laboratory confirmed cases every two years [21]. Also, the frequency of detection in older age subgroups (i.e., 2–4 years) was slightly higher than in the under 1 year old. This could be explained by the fact that the number of older susceptible children in the ongoing years since the vaccine introduction is still high.
Temporal distribution of laboratory confirmed rotavirus cases showed fluctuations. Thus, it seems difficult to determine a clear seasonal pattern because minor differences in the absolute number of cases may influence the results of this distribution given that the detection frequency of rotavirus has decreased significantly. Nonetheless, the largest amount of weekly cases was observed to be shifted to the spring season in comparison with the pre vaccination period.
Regarding molecular characterization analyses, co-circulation and annual fluctuation were observed. The genotypes G2P[4] and G3P[8], which were the most prevalent in 2016 after vaccine introduction, showed downward trend. Noteworthy, G2P[4] sharply decreased before becoming almost undetected. On the other hand, G12P[8] increased and accounted for around half of the circulating genotypes and G9P[8] has been persistently detected around 8–9% by 2018. It is important to highlight the increase in non-typeable strains reaching around 10% in 2018. These samples have been confirmed by ELISA but with our available molecular methods they could not further amplify VP7 or partial VP4 genes. Thus, some of these could have been false positive samples. Considering that viral diversity has been reported to increase in those countries with a post-vaccination setting [7], it would be interesting to study these samples with a more sensitive PCR workflow to verify results from antigen testing in the future. Also, current genotyping strategies should be updated more regularly based on the circulating strains [22]. Furthermore, if full genome genotyping cannot be performed for all the positive rotavirus samples, it would be useful to implement the genotyping of other gene segments (i.e., VP6, NSP4) to provide additional information of the genomic constellations that are circulating [23, 24].
The most outstanding findings in this study were the detection of two uncommon rotavirus strains at moderate frequencies that had circulated very sporadically or even had never been identified before. In the Americas, the G9P[4] genotype was sporadically reported in several countries, such as Brazil and the United States [8, 25]. On the other hand, Guatemala, Honduras, Mexico, Colombia and Bolivia have detected these strains at a relative frequency greater than 10% in the last decade [26–28]. However, phylogenetic analyses showed that Argentinean G9P[4] were more closely related to certain strains recently detected in Italy and India [29, 30] than to the ones circulating in our region.
The G8 genotype has been found in cattle and other species of the Artiodactyla order [31, 32]. In general, it has been highly prevalent in different countries in Africa but recently has also been detected in South Asia and Europe [33–36]. In Latin America, it has been sporadically detected in humans in association with P[4] and P[6] in Brazil, P[6] in Argentina, P[14] in Venezuela, and P[1] in Paraguay, these latter two related with interspecies transmission [37–39]. As G8P[8], it was identified in several countries of Central America and recently in Chile during a community diarrheal surveillance [40, 41]. Identity analyses of Argentinean G8P[8] detected in this study showed that they were more closely related to South Korean strains detected in 2017 [Truong et al., unpublished] than to those circulating in the neighboring Chile.
All things considered, in both cases our initial findings suggest that the emergence of G9P[4] and G8P[8] strains could be due to the introduction of new strains, rather than to reassortment events from G9P[8] or G8P[6] strains previously circulating in our country. Although more nucleotide sequences data from Latin American countries are required to understand the origin of these strains’ introduction, these analyses provide useful evidence for the hypothesis that rotavirus geographical spread dynamic is more complex than we can estimate.
While there has been a sharp decrease of G2P[4], the uncommon strains detected at moderate frequencies bore the genotype 2 genomic constellation. This observation can be explained by the depletion of population of individuals susceptible to G2P[4] after the sudden increase of this genotype in the first years after universal vaccination [14]. However, other strains that are more antigenically distant from the monovalent vaccine strain may still persist and prevail among residual cases. The high frequency of detection of these strains also suggests more efficient spreading mechanisms than other unusual G/P genotypes, and even than other G9P[4] and G8P[6] previously described. Thus, this may represent a starting point for further studies of adaptability in some uncommon strains.
Considering that the epidemiology of rotavirus has changed, it is important to support ongoing surveillance to assess whether the outcomes of this intervention remain stable, with a specific focus on the genotype diversity to add evidence for evaluation of current immunization strategies and to contribute in the development of new vaccines.