Diversity, geographical distribution, and prevalence of Entamoeba spp. in Brazil: a systematic review and meta-analysis

The genus Entamoeba includes a variety of widely distributed species adapted to live in the digestive tracts of humans and a large variety of animals of different classes. The objective of this study was to investigate the prevalence, distribution, and molecular epidemiology of Entamoeba spp. in different classes of hosts in Brazil. Studies that analyzed hosts from several classes, including humans and domestic, wild, or captive animals, were considered. The pooled prevalence of Entamoeba spp. was calculated using the random-effects model. A total of 166 studies on humans and 16 on animals were included. The prevalence of Entamoeba spp. in the Brazilian population was 22% (95% CI: 21–24). The state with the highest prevalence was Paraiba with 72%, followed by Federal District with 53%, and Rondonia with 50%. In immunocompromized patients, the prevalence was 18%, and cancer (36%) was the most prevalent cause of immunosuppression. The prevalence of Entamoeba spp. in animal hosts was 12% (95% CI: 7–17). Captive wild animals and domestic farm animals showed the highest prevalence, with 16% and 15%, respectively. The species found more often were E. coli (86.5%), E. dispar (7.9%), and E. histolytica (3.1%). In conclusion, a high prevalence (22%) of Entamoeba spp. was found in the Brazilian population, with a prevalence of up to 50% mainly in the northern, northeastern, and central-western regions. The pathogenic species E. histolytica is distributed in most Brazilian regions, with significant prevalence percentages. Among animals, unidentified Entamoeba species were most prevalent in mammals.


Introduction
The genus Entamoeba includes a variety of anaerobic, unicellular, and monoxenic protozoan species adapted to live as parasites or commensals in the digestive tracts of humans and a large variety of animals of different classes [5,7,64,110,112,205,206].
The main species of this genus that parasitize humans are E. histolytica, E. dispar, E. moshkovskii, E. coli, E. polecki, E. bangladeshi, and E. hartmanni [84,124,151,174]. Morphologically, the species E. histolytica, E. dispar, and E. moshkovskii are considered identical, but only E. histolytica is the causative agent of amebiasis, a gastrointestinal disease that commonly occurs worldwide; amebiasis is considered endemic in tropical regions and is associated with inadequate socioeconomic and sanitary conditions [8,166,216]. Entamoeba histolytica shows several degrees of virulence and is capable of invading a wide variety of tissues in the host, including those of the colon and liver, and more rarely the lung, skin, urogenital tract, brain, and spleen. This invasive feature separates it from the other species [70]. It is estimated that amebiasis accounts for 55 500 all-age deaths and causes disability-adjusted life years at 2.237 million [211].
In contrast, E. dispar can cause focal intestinal lesions in laboratory animals [133]. However, in humans, it is considered a stable commensal with no virulent characteristics, producing an asymptomatic carrier state and being generally much more prevalent worldwide than E. histolytica [64,124]. On the other hand, the idea that E. dispar is a simple commensal parasite is under discussion, and some authors discuss the importance of this species in damage of the intestine and liver [73].
Globally, the overall prevalence of Entamoeba spp. in humans is 3.5%. Entamoeba histolytica and E. dispar account for 81.7% of this global prevalence in documented infections. The comparison of prevalence by regions showed differences in prevalence between Australia (1.7%) and North America (21.6%) [64].
Regarding zoonotic potential, research on E. histolytica, E. dispar, E. hartmanni, E. coli, E. moshkovskii, and E. polecki is remarkably important because of previous reports on these species in both humans and different species of animals worldwide [76,110,152,165,206]. Furthermore, regarding pathogenic potential, some of these species can cause diarrhea and other symptomatic presentations in non-human primates [165].
In Brazil, several studies based on microscopic examination have investigated the prevalence of amebiasis in different population groups, but discriminatory studies between species (using molecular methods) are relatively scarce and mainly address different animal hosts. Although there are data on the prevalence of Entamoeba spp. in some regions, there is no aggregate analysis of the prevalence and distribution of species of this protozoan by geographic area, sex, age group, and host type in Brazil. Therefore, the objective of this systematic review and meta-analysis was to determine the prevalence and distribution of different species of Entamoeba in several host classes in Brazil.

Materials and methods
The protocol of this systematic review was registered in the International Prospective Register of Systematic Reviews (PROSPERO 2019: CRD42020167222) before its implementation. The protocol and final report were developed according to the Cochrane Handbook for Systematic Reviews of Interventions [105].

The review question
What is the prevalence and geographical distribution of Entamoeba spp. in different host species in Brazil?

Inclusion and exclusion criteria
This review included studies on various hosts (humans and domestic, wild, or captive animals) of different classes to determine the prevalence and genetic identification of Entamoeba spp. in Brazil through coprological analyses and molecular techniques.
Studies analyzing fecal samples of humans and domestic, wild, or captive animals that did not report percentages of samples positive for Entamoeba spp. were excluded. was conducted in the reference lists of dissertations and theses that evaluated the prevalence of protozoan intestinal parasites. Because this search was limited to Brazil, it was limited to studies in the English, Spanish, and Portuguese languages. This search had no start date limitation but was completed in November 2020.
The studies were searched in the following databases: Spanish Bibliographic Index of Health Sciences (IBECS), Latin American and Caribbean Literature in Health Sciences (LILACS), Virtual Health Library (BVS), US National Library of Medicine bibliographic database (Medline), Elsevier database EMBASE, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Web of Science, Cochrane Library, and National Institute of Health and Clinical Excellence (NICE). The MeSH index terms searched were Entamoeba and Brazil. The keywords Brasil and Endamoeba were also included in the search. The MeSH terms and keywords were combined via the boolean operators "AND" and/or "OR" to compose the search strings.

Assessment of methodological quality
The articles selected for data retrieval were analyzed by two independent reviewers to evaluate the methodological validity of each text before inclusion in this review. The quality of the publications included was evaluated based on the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) criteria. Studies received one point for not presenting a study design or execution limitations (risk of bias), inconsistency of results, indirectness of evidence, imprecision, and publication bias. A score of 4-5 points was considered high quality, 3 as moderate quality, and 0-2 as low quality.

Data extraction
The selected texts were evaluated by two independent reviewers for validity before inclusion; discrepancies were resolved by an independent reviewer. The data were entered into the Review Manager (RevMan 5.3) [168] software for analysis. A data extraction table was used to evaluate the quality of demographic data, study location, sample size, number of cases, number of positive cases, and diagnostic test.

Data summary
The random-effects meta-analysis model was used to analyze the overall combined prevalence of Entamoeba spp. in humans and animals. The heterogeneity among studies was evaluated using I2-statistic, which shows the percentage ofvariation among studies. These analyses were performed using the Stata software, version 12.
Regarding the methodological quality, according to the GRADE criteria used, all 166 studies evaluating the prevalence of Entamoeba spp. in different Brazilian populations as well as the 16 studies evaluating its prevalence in different animal host species presented a high methodological quality, all with a score of 5.  The prevalence of Entamoeba spp. reported in the analyzed studies was between 0.2% and 93.1%. Random-effects meta-analysis showed a pooled prevalence of 22% (95% CI: 21-24; weight 100%) of Entamoeba spp. in the Brazilian population (Fig. 2).
The analysis of pooled prevalence by state showed that it was 72% in Paraiba, 53% in the Federal District, 50% in Rondonia, 35% in Mato Grosso do Sul, 34% in Mato Grosso      (Fig. 2). The pooled prevalence with complete 95% CI values for each state is shown in Table 2.

Entamoeba spp. in animals in Brazil
The 16 studies that analyzed the prevalence of Entamoeba spp. in animals included 3805 coprological tests in different species (79.1% mammals and 20.9% birds). The classification by direct interaction with humans showed that 54% were wild animals in captivity, 2.3% were free-living wild animals, 15.2% were pets, and 28.5% were farm animals.
Of the captive wild mammals, non-human primates were the most studied, with prevalence percentages of 34% and 23%. In contrast, of the farm mammals, pigs had a prevalence of 22%. Notably, the only animal considered a pet in the studies analyzed was the dog, representing 16% (Table 4). Of the domestic farm birds, emus had a prevalence of 23% and freeliving wild birds had a prevalence of 22% (Table 4).

Entamoeba spp. diversity in different host species in Brazil
Conventional microscopy analysis, molecular characterization, serology, and isoenzyme analysis were used to identify Entamoeba spp. in 150 studies, totaling 17,651 human samples. In contrast, only six studies on host animals characterized 51 positive samples at the species level.
To calculate the prevalence of the reported species, only the samples that performed this procedure were used. For this purpose, 17,651 samples (fecal and oral cavity) with identification of Entamoeba species, were used. In these samples, the most prevalent species identified in human hosts were E. coli (86.5%), followed by E. dispar (7.9%), E. histolytica (3.1%), E. hartmanni (1.9%), and E. gingivalis 0.6% (Fig. 4). The species identified as non-pathogenic E. histolytica, through zymodeme [2,3], were considered as E. dispar. On the other hand, E. coli was the only species with a taxonomic classification, identified in animal hosts. In addition, unidentified Entamoeba species were reported in animal hosts.
The prevalence of species by geographical regions showed that E. coli was the most prevalent species in the five regions, with high percentages. Entamoeba histolytica was identified in the north (28.9%), northeast (3.4%), south (1.1%), and southeast (0.3%) regions. The southeast region presented the greatest species diversity, with the identification of the five Entamoeba spp. registered in Brazil, followed by the northeast region with four species, north and south with three, and center-west with two different species (Fig. 4).
The detailed distribution of protozoan species by the Brazilian state is shown in Figure 4.

Discussion
Data on the prevalence of Entamoeba spp. were documented in 24 of 26 Brazilian states and in the Federal District. In this meta-analysis, a pooled prevalence of 22% of Entamoeba spp. was found in the Brazilian population. The pooled prevalence was calculated with samples of studies published between 1962 to 2020, so this percentage represents an overall prevalence of Entamoeba spp. in different hosts during this period of time, in Brazil. These results reflect a sampling of the five Brazilian regions, but the northeastern, southern, and southeastern regions are better characterized since these regions present higher scientific production. The northeastern region contributed 38 articles, representing 63.3% of the samples analyzed in this meta-analysis, the southern region 27 studies (16.3%), the southeastern region 62 (12.3%), the northern region 23 (6.7%), and the central-western region 17 studies (1.4%).
The analysis of the prevalence of Entamoeba spp. by region showed contrasting realities within the states of each region. The northeastern region showed high pooled prevalence percentages in the states of Paraiba (72%), Ceara (34%), Sergipe (28%), Pernambuco (16%), Piaui (9%) and Bahia (3%). Alagoas and the Rio Grande do Norte showed another reality, with a prevalence of 4% and 2%, respectively. The central-western region showed high pooled prevalence in the Federal District (53%) and the states of Mato Grosso do Sul (35%) and Mato Grosso (34%), but the state of Goias presented a pooled prevalence of 3%. In the northern region, the states of Rondonia (50%), Para (30%), Acre (26%), Amazonas (30%) and Maranhao (19%) showed high percentages of prevalence, while and Amapa showed a prevalence of 4%. In the southeastern region, the states of Espirito Santo, Minas Gerais and Sao Paulo showed a pooled prevalences of 31%, 12% and 11% respectively, while Rio Janeiro presented a moderate prevalence of 9%. The same data were found for the southern region, where the state of Rio Grande do Sul had a high pooled prevalence of 15% and the states Parana and Santa Catarina had a moderate prevalence of 9% and 6%, respectively.
The differences in the prevalence of intestinal parasites among the Brazilian regions were recently documented in a previous study [81]. However, in addition to the differences among the regions, this present study showed great prevalence differences within the same region. This epidemiological data can be used as a tool to identify areas of social vulnerability as intestinal parasitosis is strongly associated with the Abbreviations: 95% CI, 95% confidence interval socioeconomic level of the population. In contrast, Brazil is an extensive country and presents many regional and intraregional socioeconomic and health development differences. Only 39% of the cities collect and treat 100% of the sewage [38], with the lack of adequate basic sanitation system increasing the continuous dissemination of neglected diseases linked to sanitary problems, such as intestinal parasitosis, including those caused by Entamoeba spp. Regarding sex, both showed a similar pooled prevalence of Entamoeba spp., with 29% for women and 26% for men, suggesting that sex may not be a determinant for protozoan contamination. Regarding age, there was a high prevalence in the three groups, 40% in the 10-19 years group, 34% in adults aged over 19 years, and 25% in children aged below 9 years.
Age is an important risk factor for intestinal parasitic infections. Children are often more susceptible to intestinal infectious diseases than adults owing to inadequate hygiene habits. Children aged below 9 years were the group that presented the highest number of samples analyzed in this meta-analysis, and even though it is the least prevalent for Entamoeba spp., 25% is a percentage of great importance within this population. In contrast, this study showed that the most prevalent group for Entamoeba spp. were the people aged 10-19 years. Therefore, school age represents a higher risk for amebiasis than the age of the general population. A previous study in Indonesia showed a high rate of Entamoeba spp. (52.8%) in the school-age (7-15 years) group [137]. The age group between 10 and 19 years was the most heterogeneous, including pre-adolescents, adolescents, and young adults. However, this group provides a possible panorama for the prevalence of intestinal parasitosis in high school students in Brazil.
The pooled prevalence of Entamoeba spp. infection in immunocompromized patients was 18%. This parasitic infection was most prevalent in cancer patients, with 36%, although they presented fewer samples for analysis, followed by HIV and hemodialysis patients, with a prevalence of 27% and 10%, respectively. Some studies indicate that this parasite frequently causes opportunistic infections in immunosuppressed patients [46,111]; it was one of the most common causes of morbidity in this group. This study recorded high prevalence percentages in immunosuppressed patients, especially with cancer. Cancer, HIV, and hemodialysis patients become immunocompromized as a result of the disease itself or due to therapeutic procedures that cause immunosuppression [134,193]. Although intestinal parasitic infections are a great risk with persistent diarrhea and severe clinical symptoms in immunocompromized patients, the routine diagnosis of these infections is often ignored during chemotherapy or disease [1,131]. For this reason, it is extremely important to diagnose and treat parasitic infections to decrease morbidity in this group.
The overall pooled prevalence of Entamoeba spp. in animal hosts was 12%. Of these animals, Entamoeba spp. was most prevalent in mammals (12%), followed by birds (6%). Regarding human interaction, Entamoeba spp. was most prevalent in captive wild animals, which are not easily accessible to the general population, followed by domestic farm animals. Farm animal breeding is a possible risk factor for Entamoeba spp. transmission. Therefore, it is necessary to establish control measures to minimize the transmission of these parasites among different animal hosts and humans.
For Entamoeba spp. diversity, this study showed little variability in human hosts, with differentiation into five different species. Studies on animal hosts characterized only E. coli. Of the species identified in humans, E. coli was the most prevalent (86.5%), followed by E. dispar (7.9%), E. histolytica (3.1%), E. hartmanni (1.9%), and E. gingivalis (0.6%). The prevalence of these species in Brazil determined in this metaanalysis differed from the world scenario, which presented E. dispar with the highest prevalence (49.4%), followed by E. histolytica (32.3%), E. coli (1.9%), and E. hartmanni (0.9%) [64]. The Brazilian situation could be different if the 89 studies that used conventional identification methods also used molecular analysis in the 5234 samples to separate the species E. dispar from E. histolytica, which are morphologically indistinguishable and were not included in the general percentage.
Although this study presents the commensal parasite E. coli as the most prevalent in Brazil, it is important to highlight that this species has the same transmission route as that of other pathogenic species, such as E. histolytica, E. dispar, and even Giardia lamblia as well as helminths. The prevalence of this parasite can be used as an indicator of fecal/oral transmission, suggesting intestinal parasite transmission through water supply for human consumption or through contaminated food.
Entamoeba histolytica causes severe intestinal and extraintestinal amebiasis, representing a health risk in countries with inadequate sanitary barriers. This study identified significant prevalence and distribution percentages of E. histolytica in Brazil, with 28.9% prevalence in the north, 3.4% in the northeast, 1.1% in the south, and 0.3% in the southeast. In the central-western region, no study distinguished E. histolytica from E. dispar. It is important to note that more studies need to be developed in this region to resolve this sampling bias.
This study has some limitations. First, in human studies, some authors did not distribute the positive sample results by sex and/or age, decreasing the number of classified samples to better evaluate the prevalence by these variables. Second, many samples were not identified at the protozoan species level, which could improve data on the species distribution and prevalence in Brazil, especially those of the pathogenic E. histolytica. Finally, it is recommended that publication biases be evaluated using statistical methods in meta-analyses. However, the currently available methods, such as funnel graphs and the Egger regression test, are not considered useful in proportion studies [147].
In conclusion, this study showed a high prevalence of Entamoeba spp. in the Brazilian population (22%), with a prevalence of up to 50% in the northern, northeastern, and central-western regions. Although there were contrasting prevalence percentages among the regions, there is a wide distribution of Entamoeba spp. in Brazil. There was no difference between males and females, and the age group of 10-19 years had the highest prevalence, broadly indicating the prevalence of intestinal parasitosis in high-school students in Brazil. The most diagnosed species was E. coli, which may suggest the transmission of intestinal parasites through water supply for human consumption or through contaminated food. This may lead to the possibility of infection due to other protozoan pathogenic species. The pathogenic species E. histolytica is distributed in most Brazilian regions, with significant prevalence percentages. The prevalence in mammals was the highest among animals, with interactions among humans and captive, wild, or domestic farm animals presenting the higher protozoan prevalence.
The implementation of molecular methods to detect Entamoeba spp. in scientific productions is extremely important to reduce possible false-negatives using coprological methods and to differentiate protozoan species. Patients with any type of immunosuppression should undergo routine intestinal protozoa screening and early treatment to avoid future complications because a significant prevalence was identified in this population.