Human fascioliasis in Africa: A systematic review

Fascioliasis is a globally distributed, parasitic zoonosis, caused by Fasciola hepatica and F. gigantica. A comprehensive overview of the epidemiology of human fascioliasis in Africa is missing up to now. Therefore, our objective was to conduct a systematic review aiming to summarize recent knowledge on the distribution, prevalence, and risk factors of human fascioliasis in Africa. A key word search was performed in PubMed, Web of Science and Africa Wide, to gather relevant literature, published between the 1st of January 2000 and 31st of December 2020. A total of 472 records were initially retrieved, with 40 full text articles retained for the qualitative synthesis. Human fascioliasis was reported in 12 African countries, namely Algeria, Angola, Cape Verde, Egypt, Ethiopia, Ghana, Morocco, Nigeria, Senegal, South-Africa, Tanzania and Tunisia. The majority of the studies was conducted in Egypt. A total of 28 records were population surveys. Coproscopy was the most commonly used tool for fascioliasis diagnosis in these surveys. Gender (being female), consumption of raw vegetables/seeds, age, owning livestock, and use of unsafe drinking water sources, were identified as risk factors in 7 studies. Furthermore, 43 case reports were retrieved, described in 12 studies. Eosinophilia was present in 39 of these cases, while 11 had positive coproscopy results. Eight cases described having eaten raw wild vegetables. Overall, the low number and quality of records retrieved indicates that human fascioliasis remains a truly neglected disease in Africa, and more epidemiological studies are urgently needed to both establish the actual distribution as well as risk factors on the continent.


Introduction
Fascioliasis is a globally distributed, parasitic zoonosis, caused by the liver flukes, Fasciola hepatica and F. gigantica. These parasites have a complex life cycle involving an intermediate snail host, a carrier (i.e. aquatic plants) and a final mammal host (e.g. cattle, sheep but also humans). Livestock acquire the infection when grazing on contaminated pastures, while humans typically become infected through consumption of raw water plants (e.g., watercress using a search phrase that combined search terms about humans, fascioliasis and Africa, the latter part based on the search phrase developed by Pienaar et al. [21] (for the full search phrase, and translated search phrase for each of the databases: S1 File). Moreover, reference lists of review articles were screened for relevant records, these were added as additional records.

Study selection and eligibility criteria
Datasets with retrieved records from the different databases were merged into one, after which duplicate records were removed. Then, title and abstract of retrieved records were screened for relevance. At this point, articles that focussed on non-human data, or development of diagnostic tools only as well as review articles, were removed. Subsequently, full texts of the remaining records were assessed for eligibility. Exclusion criteria were: (i) studies concerning a different parasite than F. gigantica or F. hepatica, ii) studies on fascioliasis in animals or presence of the parasite in the environment, ii) studies published before 2000 or after December 31st 2020, iii) studies reporting results from outside the study area, iv) studies reporting results out of the scope of the review question, v) duplicate records. For the third criterion, case reports for people of non-African origin diagnosed outside Africa were excluded, yet case reports for people of African origin diagnosed outside Africa area within 6 months of arrival (and with country of origin mentioned), were included. The PRISMA guidelines were followed for reporting the review [22] (Fig 1, S2 File).

Data collection process, data items and evaluation study quality
Retrieved articles were classified as either population survey or case report, and the following variables were collected: i) population surveys: population studied, study period, population setting, number of people tested, number of positive individuals, prevalence, diagnostic tests used, risk factors associated with the disease (e.g. odds ratios), ii) case reports: study period, gender, age, clinical signs and symptoms, diagnostic tests used and test results. In case population surveys reported findings of interventions, only baseline pre-intervention data were extracted. Furthermore, publication year was extracted and it was checked whether publications were published in a journal listed with an Impact Factor (IF) in the Science Citation Index (SCI). All data were entered in preformatted tables. Next, the study quality of population surveys was assessed by the Joanna Briggs Institute Prevalence Critical Appraisal Tool [23]. After the evaluation, the number of questions scoring a "Yes" was calculated, and divided by the total number of questions, the latter not including questions which were deemed not applicable for the study. Studies scoring less than 50%, were labelled of "weak quality", between 50 and 75% of "moderate quality" and equal to or above 75% of "strong quality".

Summary measures and synthesis of results
For the population screening studies, a descriptive statistical analysis was undertaken whereby the proportion of people infected was calculated based on the number of people positive for fascioliasis and the number of people tested. The associated Wilson score 95% confidence intervals were calculated for these proportions. Chi-square tests were run to investigate the association between risk factors and presence of disease; in case of cell counts below 5, Fisher exact tests were conducted instead. Odds ratios for the risk factors were calculated as well as associated Wilson score 95% confidence intervals. The significance was set at the 5% level. All statistical analyses were carried out using R, version 3.6.1 [24].

Results
A total of 467 publications were extracted from the three scientific databases, and 5 additional records were identified through reference list screening (Fig 1). After duplicate removal, titles and abstracts of 307 remaining records were screened for relevance. Then, the full text articles of the remaining 68 records were assessed for eligibility. Twenty-eight records were excluded at this stage: 13 were excluded due to the study topic being out of scope for the current systematic review, 6 did not focus on the study area, 2 contained data on a parasite different from Fasciola spp., 1 record reported duplicate results. For 6 records, the full text could not be retrieved. Finally, 40 records were retained for the qualitative synthesis.
A total of 28 records presented data from 36 population surveys conducted in Egypt (n = 15), Nigeria (n = 7), Angola (n = 1), Ethiopia (n = 1), Ghana (n = 1), South Africa (n = 1), Tanzania (n = 1), and Tunisia (n = 1) (Table 1), with common groups studied being the general population, schoolchildren and patients attending hospitals. The prevalence in the general population, and healthy adult and children subgroups varied widely, ranging between 0.29 and 19.3%. The most frequently used diagnostic tool was coproscopy, with only one survey providing morphometric details used for Fasciola spp. identification [25]. Few studies used serological methods (e.g. hemagglutination test, ELISA) [26,27] or imaging as diagnostic tool [28]. Six studies described co-infection, mainly with Schistosoma spp. in surveys specifically targeting co-infections between those species [29][30][31][32][33][34][35]. None of the population surveys performed speciation (i.e. determination whether Fasciola hepatica or F. gigantica caused the infection). The population survey quality evaluation indicated that, for a good number of studies, inadequate attention was given to sample size calculation (23/28), measures to address non-responders (22/27), reliability of outcome measurement (20/28) and subject selection process (10/28). Twelve studies were scored as weak, 12 as moderate and only 4 as of strong quality (S3 File).
Seven population surveys investigated risk factors associated with fascioliasis ( Table 2) in Egypt (n = 5), Nigeria (n = 1), and Ethiopia (n = 1), all but one [27] using coproscopy for identification of positive cases. Gender (being female) and consumption of raw vegetables/seeds were identified as risk factors in 2 studies each. Other risk factors identified in the 7 studies were age, owning livestock, and use of unsafe drinking water sources.
Twelve case reports were retrieved for Egypt (n = 2), Ethiopia (n = 2), Morocco (n = 2), Tunisia (n = 2), Algeria (n = 1), Cape Verde (n = 1), Senegal (n = 1), and South Africa (n = 1). A total of 43 cases were described (Table 3), with an average age of 28 years old and 22 out of 43 cases being male. Fifteen cases mentioned some sort of abdominal pain, with 7 indicating that they suffered from epigastric pain specifically. Another 15 cases described fever, while 4 reported that they did not suffer from any clinical signs. For all but 3 out of 42 cases, for whom the information was available, eosinophilia was present, with percentages of eosinophils up to 80% [36]. Eight cases described eating raw wild vegetables. Out of 39 cases with coproscopy results available, Fasciola eggs were detected in 11. Only 5 studies provided some morphometric evidence for Fasciola spp. Infection, of which 2 provided pictures of eggs [37,38], and 3 of Fasciola spp. Adults [39][40][41]. Other tests used to establish infection, were Ab-ELISA [42], electrophoresis [36,43], hemagglutination [43] and indirect fluorescent antibody tests [44]. None of the case reports performed speciation (i.e. determination whether Fasciola hepatica or F. gigantica caused the infection).

Discussion
Human fascioliasis was found to occur in a number of countries, spread throughout the African continent. The majority of studies included in the review were conducted in Egypt, others were conducted in Algeria, Angola, Cape Verde, Ethiopia, Ghana, Morocco, Nigeria, Senegal,  where F. hepatica is the prevailing species [13,14]. More studies on human fascioliasis should be conducted in the remaining countries in Africa, for which we could not retrieve records, and where possible, with the speciation of the causative parasite (i.e. F. hepatica or F. gigantica).   In most of the countries where population surveys were conducted, 1-2% of the screened population were found to be positive for fascioliasis. Globally, the epidemiological scenarios for fascioliasis are rather heterogeneous, ranging between zones with only imported cases, and hyperendemic areas, with prevalence estimates over 10% [45]. According to the classification proposed by Mas-Coma, Valero and Bargues [45], most of the countries for which records were retrieved would be considered meso-endemic areas, although others would rather be considered hypo-or hyperendemic. Moreover, it is well known that the prevalence of human fascioliasis does not correlate well with the prevalence of fascioliasis in livestock, the latter also generally being higher than the former [46]. This seemed also true for the prevalence of human fascioliasis in Africa, as hyperendemic areas for human fascioliasis did not always overlap with regions with high prevalence estimates in livestock, and prevalence estimates were usually lower (1-2%) than those generally reported for livestock in the region (i.e. above 10%, even above 50% at times) [15,16]. Overall, more attention should be given to further identification of hyperendemic areas for human fascioliasis on the continent. Also, intensity of infection, measured as eggs per gram of faeces (EPG) should be reported considering that intensity is linked to pathogenicity of infection [47], and in view of the high intensities recently reported in Egypt [48].
Coproscopy was the most commonly used diagnostic tool in the retrieved population surveys. This choice might have had an impact on the prevalence estimation in the surveys. Indeed, many cases might have been missed, due to the poor sensitivity of the test, the inability to detect fascioliasis at an early stage, and the very low egg shedding, especially in low infection burdens as well as in old infections [49]. On the other hand, coproscopy can give false positive results due to spurious infection following the consumption of livers or guts, contaminated with Fasciola or Paramphistomum, and misdiagnosis of eggs from other trematodes (e.g. Gastrodiscoides hominis, Paragonimus spp.) [6]. WHO has proposed a combined evaluation of test results (e.g. serology, coproscopy) to allow differentiation of different stages of the infection, yet interpretation remains challenging [1]. Ultrasound examination might be considered another additional tool to detect fascioliasis related lesions. However, ultrasound is not always able to differentiate fascioliasis from other liver diseases because fascioliasis causes unspecific lesions and it is often unavailable in a resource-poor setting [49].
Despite the importance of the disease on the continent, the number of reports on the topic has been consistently low over the study period. The situation did not seem to improve after 2010, the year fascioliasis was listed a neglected tropical disease by the WHO [12]. Moreover, a noted discrepancy was found between the countries having reported human fascioliasis cases that were retrieved in our study as compared to the WHO fascioliasis distribution map [20], with countries present on our distribution map being absent on the WHO map (e.g. Tanzania, Angola, South Africa), and vice versa (e.g. Mali, Niger, Cameroon), although the WHO map did include information from prior 2000, in contrast to our review. This all points to limited attention for and thus knowledge on the occurrence of human fascioliasis on the African continent.
Moreover, there exists a limited understanding of the risk factors for fascioliasis in the African setting. Our search retrieved only 7 studies investigating risk factors, most of which were conducted in Egypt. In the retrieved records, consumption of raw vegetable/seed consumption and use of unsafe drinking water, both established routes of infection for fascioliasis worldwide, were significantly associated with fascioliasis [19]. Moreover, being female was found to be significantly associated with fascioliasis [35]. This could be attributed to the traditional gender roles in Africa, where women and girls are more involved in the preparation of meals, washing clothes and kitchen utensils with contaminated water than their male counterparts [30]. The background of other identified risk factors such as, owning livestock and performing irrigation practices [29] might be related to contact with contaminated water.
Apart from the risk factors identified in the studies retrieved in this systematic review, other behavioural factors might favour transmission of fascioliasis, but remain largely unstudied. For instance, authors have named traditional beverages made from sylvatic vegetables and sugarcane grown in swampy areas as source of infection in Cape Verde [19,50]. In the Horn of Africa, chewing khat (Catha edulis) was linked to several fascioliasis cases in travellers [19,50]. Other potential sources of infection might be cabbage, or other vegetables consumed uncooked that are grown in swampy areas, grass or sugar chewing [19]. Up to now, however, none of these factors have been investigated in large-scale epidemiological studies.
Our study has some limitations. First, we might have missed data, as we did not have access to potentially relevant grey literature, the amount of which might be considerable in Africa. Secondly, most retained studies, including those investigating risk factors, used coproscopy as main diagnostic test. Due to the low sensitivity and specificity of coproscopy for the diagnosis of fascioliasis [49], disease prevalence estimates might be misleading. Moreover, certain risk factors might have been missed due to the impact of the imperfect test on odds ratio estimation [51]. Nevertheless, this is the first study gathering available information on human fascioliasis epidemiology in Africa.
Overall, the low number of records retrieved indicates that human fascioliasis remains a truly neglected disease in Africa. Apart from the need for more appropriately performed screening studies to estimate its prevalence in all countries on the continent (considering its worldwide distribution), an in-depth investigation of local risk factors is lacking yet paramount to fight fascioliasis in Africa.