Systematic Review of Hepatitis E Virus in Brazil: A One-Health Approach of the Human-Animal-Environment Triad

Simple Summary Hepatitis E virus (HEV) is an important causative agent of acute and chronic hepatitis worldwide. Originally identified in epidemics associated with flooding in Asia, it nowadays shows very distinct genetic and epidemiological patterns. While HEV genotypes (HEV-) 1 and 2 are associated with the original outbreaks (waterborne diseases), HEV-3 and HEV-4 present a zoonotic pattern (associated with consumption of meat from infected animals), HEV-5 and 6 have been found only in wild boar in Japan, and HEV-7 and 8 have been detected in camels and dromedary seldom affecting humans. Brazil, with a precarious sanitary structure and being an important world meat producer, was the focus of this study in order to identify patterns of occurrence of HEV. After reviewing scientific studies, it was identified that the only genotype found in Brazil is HEV-3 and the area where there were more reports was the South region of the country. This is the region that produces more pork. These results indicate that HEV-3 is widespread in the country and sanitary surveillance is essential in the national production of pigs, as well as the implementation of monitoring protocols in hospitals. Abstract Brazil is the fifth largest country in the world with diverse socioeconomic and sanitary conditions, also being the fourth largest pig producer in the world. The aim of the present systematic review was to collect and summarize all HEV published data from Brazil (from 1995 to October 2020) performed in humans, animals, and the environment, in a One Health perspective. A total of 2173 papers were retrieved from five search databases (LILACs, Mendeley, PubMed, Scopus, and Web of Science) resulting in 71 eligible papers after application of exclusion/inclusion criteria. Data shows that HEV genotype 3 (HEV-3) was the only retrieved genotype in humans, animals, and environment in Brazil. The South region showed the highest human seroprevalence and also the highest pig density and industry, suggesting a zoonotic link. HEV-1 and 2 were not detected in Brazil, despite the low sanitary conditions of some regions. From the present review we infer that HEV epidemiology in Brazil is similar to that of industrialized countries (only HEV-3, swine reservoirs, no waterborne transmission, no association with low sanitary conditions). Hence, we alert for the implementation of HEV surveillance systems in swine and for the consideration of HEV in the diagnostic routine of acute and chronic hepatitis in humans.


Introduction
In the last years, hepatitis E virus (HEV) has captured widespread attention when autochthonous hepatitis E cases started to be reported in industrialized countries [1]. Until screened the databases, and relevant information was extracted. Differences in opinions about whether to include an article were solved by consensus between all the authors.

Results
A total of 2173 papers were retrieved from the 5 databases used for the search (Figure 1). After removal of duplicated papers (n = 542), exclusion criteria were applied to eliminate non-related papers, namely papers classified as "non-Brazilian" (n = 24), "non-HEV" (n = 1519), as well as review articles and in vivo animal experimental studies. Application of inclusion and exclusion criteria generated a total of 71 eligible papers. They were all included in the study after being assessed by full-reading. The distribution of published papers by regions of Brazil and type of study can be observed in Figure 2. HEV studies in humans, swine and animal products, animals other than swine, and environment are summarized in Tables 1-4, respectively. The RNA load ranged from 10 2 -10 8 copies/mL [32]  The RNA load ranged from 2500-4000 copies/mL  Table 2. HEV in swine and animal products, Brazil.

HEV in Humans
HEV studies performed in humans in Brazil (Table 1) were focused on a variety of population groups and most were serological surveys.
Studies performed in populations from regions with lower sanitation and hygiene conditions in the North region found an anti-HEV IgG seroprevalence of 0.3% in afro descendants [14]. Studies done in poor communities in the Midwest region found an anti-HEV IgG seroprevalence of 3.3% and 10.66% in adults [28,30] and 4.5% in children [27]. In the Southeast region, a seroprevalence of 2.4% was found also in poor communities [34].
Seroprevalence studies focusing on rural settlements (Table 1) found anti-HEV IgG seroprevalences of 12.9% in the North [11], 3.4% [25], 3.9% [23], and 8.4% [26] in the Midwest, and 2.1% [35] and 20.7% [41] in the Southeast. Three of these studies performed in rural settlements were also focused on current and/or recent infections. The study of the Midwest region found 0% of anti-HEV IgM and HEV RNA [25] and the study of the North found 0.3% of anti-HEV IgM [14].
HEV studies in Brazil have also focused on transplant recipients (Table 1). Among those with kidney transplants, anti-HEV IgG seroprevalence was found to be of 2.5% [21] in the Midwest, and 3.1% [45] and 15% [43] in the Southeast. HEV RNA was found in 3.1% [45] and 10% [43] of kidney transplant recipients. Only two studies investigated HEV infection in liver transplant recipients, namely a case report in a pediatric patient [40] and a study in the Southeast region that found a seroprevalence of anti-HEV IgG and IgM of 8.1% and 2.6%, respectively [39].

HEV in Swine and in Animal Products
All studies performed in swine ( Table 2) found evidence of HEV infection, either by using the detection of anti-HEV IgG and/or HEV RNA. Seroprevalence studies in younger pigs (<10 months) found an anti-HEV IgG prevalence of 8.6% in North region of Brazil [57] and 69.7% in the Midwest region [60]. The detection of HEV RNA in stools in this age group was 1.7% in the Northeast region [58] 7.94% in the North region [57] and 87.5% in Southwest [67].
In pigs from family-scale the anti-HEV IgG prevalence was 0% [61] and 67% [60] in the Midwest region, and 77.6% in the South region [71]. Regarding the detection of HEV RNA in stools of pigs from family-scale farms, 8% [61] and 24% [62] were found positive in the Midwest region, and 20% [68] in the South region.
Seroprevalence studies on slaughtered pigs showed anti-HEV IgG in 81.2% in Midwest [64] and 81.3% in the Northeast [59]. The detection of HEV RNA in bile from slaughtered pigs showed to be positive in 9.6% [66] and 15.2% [65] in Southeast and 0.84% in South [69].
Concerning wild boar, only two HEV seroprevalence studies were performed, both in the South region, having found a seroprevalence of 14.29% in Rio Grande do Sul state [73] while in Santa Catarina state, 1.55% [73] and 13.1% [74] seroprevalences were observed.
Regarding the HEV contamination of meat and meat products derived from swine and other animals ( Table 2), HEV RNA was detected in 36% of the pig pâtés and blood sausages (morcilla) derived from pork [76]. In another study, no HEV was detected either in pig processed meats such as mortadella, sausage, salami, ham, and pate, or in the raw meat of bovine, swine, chicken, and capybara [75].

Discussion
The HEV studies in humans in Brazil started in the early 90s. The majority of these initial investigations were conducted in rural areas, possibly motivated by the HEV-1 and HEV-2 data from endemic regions in developing countries with similar poor sanitary conditions. The first HEV reports in Brazil focused on communities with low levels of sanitation, such as gold miners [29] and poor communities [28,30] from the Amazon area of the Midwest region, and from the Southeast region [34]. In these reports, the fecally contaminated water was pointed as a potential route of HEV transmission and the seroprevalences within these communities ranged from 0.45% in children to 10.66% in adults [27,28].
After the recognition of HEV-3 as being responsible for autochthonous hepatitis E in industrialized countries [81,82], HEV studies in Brazil started to focus on cases of acute non-A-C viral hepatitis in order to clarify the potential role of HEV in these undiagnosed cases [17,28,35], efforts that still motivate publications nowadays [15,36]. In general, markers of current and/or recent HEV infection (anti-IgM HEV and HEV RNA) have been detected but at a low prevalence, indicating that HEV was not the causal agent of the majority of these acute hepatitis cases.
Based on the knowledge that HEV-3 infection may progress to a chronic hepatitis in immunocompromised patients [3], some HEV studies in Brazil have focused on organ transplant recipients [39] and HIV patients [42]. In kidney transplants, HEV seroprevalence varied from infrequent (2.5%) [21] to frequent (15%) [43]. In liver transplant recipients the prevalence of anti-HEV antibodies showed to be higher than immunocompetent populations in Brazil, suggesting HEV infection as a possible cause of liver injury [39]. Concerning HIV patients, studies showed similar HEV seroprevalences when compared with blood donors indicating that HIV patients are not at risk for HEV infection [19,53].
Hepatitis E caused by HEV-1 and HEV-2 has been associated with morbidity and mortality in pregnant women [3]. Possibly motivated by this, some HEV seroprevalence studies have been performed in pregnant women in Brazil, however no risk for HEV seropositivity has been shown in this particular group when compared with the general population [13,35,49].
Several studies have evaluated the HEV seroprevalence in the general population of Brazil, with the majority using blood donors as the sampled group. A great range of HEV seroprevalence was observed, with the lowest detected in the North (0.45%) [13] and Northeast regions (2%) [18]. Mid-range levels of HEV seroprevalence were observed in the Midwest (4%) and Southeast (4%, 9.8%) regions [26,37,47]. In the South region, the five seroprevalence studies showed values of 2.3% [50], 7.1% [53], 10% [54], 26% [49], and 40.25% [52]. The high seroprevalence detected in the South has been justified for being the region in Brazil with the highest density of pig farms and the largest consumption of pig meat and related products [52]. In fact, pig breeding has been suggested to influence human HEV seroprevalence in other countries [83,84]. Epidemiologic surveys performed in rural population of Brazil, namely in the North [11] and in the Southeast regions, have found higher seroprevalences in these populations (12.9% and 20.7%, respectively) when compared to those previously reported on blood donors from the same regions [11,41]. This difference has been attributed to the lower sanitary conditions of the rural populations. Overall, the range of seroprevalences observed in Brazil has to be interpreted with caution since some studies were performed several decades apart and using different immunoassays. It is widely known that the different anti-HEV IgG immunoassays and their performance characteristics strongly influence HEV seroprevalence data [85].
Despite the strong evidence of widespread HEV circulation in Brazil, the recent report of the official governmental databases presented no notification of hepatitis E among the notified 216,379 hepatitis cases [86]. This draws attention to an underdiagnosis and/or underreporting of hepatitis E in Brazil. The underdiagnosing of hepatitis E cases has been reported elsewhere and is partly attributed to the fact that HEV testing has not been traditionally included in hepatitis differential diagnostic algorithms [87].
Many HEV studies in Brazil have focused on swine, which is understandable given the fact that this country is the 4th largest pig producer in the world, with more than 2 million breeders and producing 3975 thousand tons/year of pork meat, with the South region representing 66.12% of the national production [88]. Circulation of HEV in pigs of Brazil was observed either in large or family-scale herds, and in all age groups, based on HEV RNA presence in stools/biological fluids/organs (0.8-88.9%) or anti-HEV IgG seroprevalence (0-77.6%) [61,62,68,72]. Evidence for HEV infection in slaughtered pigs was also shown by the high seroprevalence (>80%) detected [59,64]. The circulation of HEV was also demonstrated in wild boars of Brazil with seroprevalences ranging from 1.55% to 14.29% [73,74]. HEV was inclusively found in pig pâtés and blood sausages derived from pork [76]. Overall, HEV is highly disseminated in the swine population throughout Brazil and might present a risk to animal handlers and pork consumers, mainly if pork meat and meat products are eaten raw or undercooked. The presence of HEV in pigs and derived pig products has been widely reported in other countries [84,[88][89][90].
In the past years there has been an interest in studying HEV infection in non-human primates, inclusively Macaca fascicularis were used on experimental in vivo studies performed in Brazil to evaluate HEV pathogenesis [91][92][93]. HEV seroprevalences have been reported in farmed Rhesus monkeys in China (70.8%) [94] and in captive non-human primates in Italy (4.2%) [95] but the only seroprevalence study performed in Brazil in wild non-human primates did not detect any (0%) anti-HEV antibodies [11]. Furthermore, no HEV RNA was detected in the stools and livers of Golden-headed lion tamarins of Brazil [77].
Serological studies in Brazil also focused on other animals, having reported the presence of antibodies anti-HEV in cows, dogs, chicken, and wild rodents, but not in sheep and goats [11]. Antibodies against HEV have also been detected in dogs in the United Kingdom [96], in chicken, cows, wild rodents, sheep, and goats in China [97][98][99][100], chickens in Korea [101], sheep in Italy [102], but the zoonotic importance of these animals concerning HEV remain to be clarified. Noteworthy, two novel HEV strains were discovered in wild rodents from Brazil (Calomys tener and Necromys asiurus) [78].
Concerning the HEV studies that focused on the environment in Brazil, only water samples under the influence of swine farm effluents, namely slurry lagoons, were found positive for HEV [66,72]. Samples from the southern region of Brazil, with a high density of swine production, detected HEV in up to 100% of the samples analyzed [72]. This same region coincides with the highest rates of human seropositivity for HEV and is also the region with the highest concentration of pig production in the country. This fact, analyzed from the One Health perspective, highlights the zoonotic character of this virus. Swineinfluenced waters contaminated by HEV have been frequently detected and reported in other countries [103,104]. In the studies of Brazil, HEV was not detected in bivalve molluscs, recreation waters, or even in waters that drained effluents from pig farms or waters of poor quality, very close to human settlements [76,79,80]. However, studies in other countries have reported HEV in bivalve molluscs [105][106][107], seawater [108], and wastewater [109,110]. These discrepancies of detection of HEV in environment samples could be in part due to the low concentration of HEV and complexity of the matrices, two well-known limiting factors of the detection of enteric viruses in environmental samples.
Concerning the molecular characterization of HEV strains detected in Brazil, studies showed that all HEVs found in Brazil were classified as HEV-3 (6 studies in humans, 15 in swine and animal products, and 2 on environmental samples). HEV-3 is known to have a zoonotic (swine) origin and the subtypes 3b and 3i were detected in humans [33,40,45] and pigs [61,62,65,66,[68][69][70][71][72], while the subtypes 3c [57,65,71], subtype 3d [61], subtype 3f [57,58,62] and subtypes 3h [61,71] have been only detected in pigs. As molecular studies have been performed using several molecular assays and primer choices, different regions of HEV have been targeted and characterized. This clearly hampers the robust classification of HEV subtypes and, consequently, a solid comparison between subtypes, hence caution must be taken when analyzing this data. In fact, attention should be paid to several factors that could bias the interpretation of results here presented. A clear focus has been given to human samples with little attention to animal or environmental matrices, most likely due to the initial understanding of this disease, not known to be zoonotic at that time. Additionally, not only a higher number of studies have also focused on the South where the highest density of pig farms is present but also a vast diversity of sample sizes has been used throughout the studies, making it difficult to robustly compare results. Further studies spatially dispersed are for these reasons recommended.
The present systematic review is not the first that targets HEV in Brazil. The two published so far have centered only on human infection [111,112] while here we present for the first time a perspective focusing on the One Health triad, having included HEV studies on humans, animals, and environment. A One Health approach makes it possible to look at issues such as zoonotic diseases, food safety, and food security, as well as environmental contamination and other aspects. In this perspective this review evidenced that the scientific community has approached the topic of HEV on every aspect of environment, human, and animal systems individually, however when compiled, this translates into data that broadens the scope to One Health.

Conclusions
Overall, this systematic review shows that HEV-3 was the only retrieved genotype in humans, animals, and environment in Brazil. The South region showed the highest HEV seroprevalence in humans, which curiously is also the region with the highest pig density, swine industry, and pig HEV circulation, suggesting a zoonotic link. HEV-1 and HEV-2 were not detected in any of the studies performed in Brazil, even in those focusing on low sanitary condition communities. This allowed us to infer that HEV epidemiology in Brazil is similar to that of industrialized countries (only HEV-3 circulation, swine reservoirs, no waterborne transmission, no association with low sanitary conditions). Hence, we alert for the implementation of HEV surveillance systems in swine and for the inclusion of HEV in the diagnostic routine of acute and chronic hepatitis in humans. More sequence data are needed on HEV strains circulating in humans, animals, and the environment to further evidence the zoonotic origin of HEV infection in Brazil.